search this blog

Sunday, May 19, 2019

Who were the people of the Nordic Bronze Age?


Ancient DNA has revealed that large scale migrations and population replacements have often accompanied major cultural changes in prehistoric Europe. But, for now, my opinion is that the formation of the archeologically ostentatious Nordic Bronze Age wasn't associated with any significant foreign gene flow into Scandinavia. I've tested this as best as I could with the few relevant ancient samples that are currently available.


For instance, below are among the most successful qpAdm mixture models that I was able find for various ancient Scandinavian groups dating back to the local Middle Neolithic (MN) period. The Nordic Bronze Age population is represented by three individuals labeled Nordic_BA. Unfortunately, the guy pictured above, from the famous Borum Eshøj barrow burial in what is now Denmark, didn't make the cut. For more details about my sampling and labeling strategies refer to the text file here.

Nordic_MN_B
CWC_CZE 0.822±0.059
POL_Globular_Amphora 0.178±0.059
chisq 14.478
tail prob 0.341086
Full output

SWE_Battle_Axe
CWC_Baltic_early 0.662±0.028
POL_Globular_Amphora 0.338±0.028
chisq 11.234
tail prob 0.591189
Full output

Nordic_LN
Nordic_MN_B 0.928±0.069
SWE_TRB 0.072±0.069
chisq 12.139
tail prob 0.516307
Full output

Nordic_BA
Nordic_LN 0.851±0.061
SWE_TRB 0.149±0.061
chisq 10.897
tail prob 0.619475
Full output

It's impossible to successfully model the ancestries of Nordic_MN_B and SWE_Battle_Axe simply with the populations that were living in Scandinavia before them. Therefore, it's likely that they were migrants or the recent descendants of migrants to Scandinavia. But there's nothing surprising about that, because they're archeologically associated with the Corded Ware culture (CWC), which has always been seen as intrusive to Scandinavia from the south and east.

Conversely, it's easy to produce statistically sound mixture models for both Nordic_LN and Nordic_BA exclusively with earlier Scandinavian populations. Indeed, based on the outgroups or right pops that I'm using, Nordic_LN is almost indistinguishable from Nordic_MN_B, and the same can be said of Nordic_BA in regards to Nordic_LN.

Of course, if I mixed and matched reference populations from across prehistoric Europe, I could probably come up with some spectacular statistical fits even without the need for any Scandinavians. Essentially that's because Nordic_LN and Nordic_BA are closely related to many earlier and contemporaneous peoples living all the way from the Atlantic facade to the Ural Mountains. My point, however, is that this isn't crucial, despite the dearth of ancient samples from Scandinavia.

This is how things look in a Principal Component Analysis (PCA) of Northern European genetic variation based on my Global25 data. Strikingly, Nordic_MN_B, SWE_Battle_Axe, Nordic_LN and Nordic_BA more or less recapitulate the cluster made up of present-day Swedish samples. The relevant datasheet is available here.
Granted, two of the Nordic_BA samples sit just south of the Swedes, no doubt due to their slightly higher ratios of Neolithic farmer (SWE_TRB-related) ancestry, but this is also an area of the plot that many present-day Danes call home (not shown, because I don't have any suitable academic Danish samples to run).

I'll eat my hat if it turns out that Scandinavia experienced a major population shift (say, more than a collateral ~10%) during the LN and/or BA periods. And I'll post a clip of it online too.

Update 27/08/2019: Four of the samples from the recent Frei et al. paper on human mobility in prehistoric southern Scandinavia are in my Global25 datasheets. So I thought it might be interesting to check whether their strontium isotope ratios correlated with their genomic profiles.

In the Principal Component Analysis (PCA) below, RISE61 is a subtle outlier along the horizontal axis compared to the other three Nordic ancients, as well as a Danish individual representative of the present-day Danish gene pool. Also note that RISE61 shows the most unusual strontium isotope ratio (0.712588). The PCA was run with an online tool freely available here.


To help drive the point home, here's a figure from Frei et al., edited by me to show the positions of RISE47, RISE61 and RISE71. If RISE276 was also in this graph, he'd be sitting well under the "local" baseline, in roughly the same spot along the vertical axis as RISE47.


Interestingly, RISE61 belongs to Y-chromosome haplogroup R1a-M417, while RISE47 and RISE276, who appear to have been locals, both belong to R1b-M269. My guess is that RISE61 was a recent migrant from a more northerly part of Scandinavia dominated by the Battle-Axe culture (BAC). The BAC population was probably rich in R1a-M417 because it moved into Scandinavia from the Pontic-Caspian steppe via the East Baltic. This is what Frei et al. say about RISE61 and his burial site:

The double passage grave of Kyndeløse (Fig 1, S1 File) located on the island of Zealand yielded 70 individuals as well as a large number of grave goods, including flint artefacts, ceramics, and tooth and amber beads. We conducted strontium isotope analyses of seven individuals from Kyndeløse encompassing a period of c. 1000 years, indicating the prolonged use of this passage grave. The oldest of the seven individuals is a female (RISE 65) from whom we measured a “local” strontium isotope signature ( 87 Sr/ 86 Sr = 0.7099). Similar values were measured in five other individuals, including adult males and females. Only a single individual from Kyndeløse, an adult male (RISE 61) yielded a somewhat different strontium isotope signature of 87 Sr/ 86 Sr = 0.7126 which seems to indicate a non-local provenance. The skull of this male individual revealed healed porosities in the eye orbits, cribra orbitalia, a condition which is possibly linked to a vitamin deficiency during childhood, such as iron deficiency.

By the way, RISE47 was buried in a flat grave, which suggests that he was a commoner. RISE276 was found in a peat bog in Trundholm, where the famous Trundholm sun chariot was discovered (see here). He may have been a human sacrifice.

Citation...

Frei KM, Bergerbrant S, Sjögren K-G, Jørkov ML, Lynnerup N, Harvig L, et al. (2019) Mapping human mobility during the third and second millennia BC in present-day Denmark. PLoS ONE 14(8): e0219850. https://doi.org/10.1371/journal.pone.0219850

See also...

They came, they saw, and they mixed

Children of the Divine Twins

The mystery of the Sintashta people

258 comments:

1 – 200 of 258   Newer›   Newest»
Synome said...

Interesting implications if e.g. Bell Beaker did not make an appreciable genetic impact in the NBA. Since linguistic evidence suggests a heavy Italo-Celtic like influence in Proto-Germanic, this raises the question of how this influence occurred.

Was the ancestor of Proto-Germanic already like this in the Chalcolithic?

If it acquired the influence later, where is the genetic impact form a candidate IC culture like Beaker?

Ric Hern said...

So basically it indirectly points towards a very early split of Germanic from its relatives, Celtic and Italic. Makes sense since Celtic and Italic seems a little closer to each other...and it could point to a much older date of origin of the Indo-European Language Families of Western Europe...

John Thomas said...

Everyone is Polish. Only they don't know it : )

AWood said...

R1a-Z284 and L664 appear to be related to CWC in some fashion, or descending from them. The latter possibly being absorbed by Beaker groups, and the former as Scandinavian offshoots from the Battle Axe period. The only R1b linked to this period that is Scandinavian specific is R1b-L238, which is an early node under P312+, and might be have travelled with the R1a-Z284 group. That said, this would infer that L11+ was probably not far, occupying a large swath of territory in modern day Germany based on the success rates of his descendants U106+ and P312+. I'd argue much of the R1b moved north in the period following CWC/Battle Axe, which would frame things around the Nordic BA. The missing link is really I1, since it's all over the place in northern Europe. It's a bit of a repost from something earlier, but it's more relevant here.

Andrzejewski said...

I1 I believe is represented in Swe_TRB?

Angantyr said...

@Andrzejewski

It's better to check facts than to believe. The first I1 is RISE179, from Nordic LN II (that is, the late late Neolithic - which is basically the same as the early Bronze Age).

weure said...

Wow fascinating, this continuity is not only the case in Scandinavia but also in the Northern Netherlands. I plot right on Rise71, Bronze Age Denmark. I find this kind of remarkable continuity from the Bronze Age until now!?

I get bugs with this datasheet (my fault!!). But previous year I was able to plot myself, I transponded this into the new figure:

https://www.mupload.nl/img/u1dk6rrhw9.37.59.png

I'm the red dot.

So is this the (proto) Germanic continuity and genetic profile?

weure said...

This makes the Davidski theory more plausible le. Single Grave people went to Denmark and North Dutch and transformed not only into BB but (afterwards) also mingled with the previous TRB people (TRB had also hotspot in Denmark and Northern Netherlands).

So no wonder that not only the Danes but also North Dutch like me plot on the same spot.

Chapeau Davidski!

rozenfag said...

Question: are there any news about Tollense guys?

Arza said...

@ rozenfag

https://www.jugend-forscht.de/fileadmin/user_upload/Downloadcenter/Bundeswettbewerb/L%C3%A4nderlisten_Projekte/Projekte_Hamburg_2019.pdf

Stand 23
Biologie
Kerrin Bielser (19)
Hamburg
Universität Rostock
Erarbeitungsort: Institut für Bioinformatik, Université de Fribourg, Schweiz
Programm zum Schätzen des Inzuchtkoeffizienten

Im Tal des Flusses Tollense in Mecklenburg-Vorpommern wurden seit den 1990er Jahren Knochen gefunden, die auf eine blutige Schlacht in der Bronzezeit hinweisen. Von einigen der Getöteten konnte Genmaterial sichergestellt werden. Kerrin Bielser analysierte die vorhandenen DNA-Daten, um den Grad der Inzucht unter den beteiligten Personen festzustellen. Zu diesem Zweck schrieb sie ein Programm in der Sprache R, das umfangreiche Statistiken nutzt. Die Berechnungen offenbarten, dass die Krieger nur in geringem Maße verwandt waren. Daraus schließt die Jungforscherin, dass die Krieger aus unterschiedlichen Regionen stammten und sich für die Kämpfe offenbar organisiert hatten. Allerdings war die Anzahl der untersuchten Individuen noch relativ gering, sodass die Aussage durch weitere Analysen gefestigt werden müsste.



They are using truncated data for meaningless side-projects only. I don't think that we will see full DNA sequences anytime soon. German museums currently claim that no one knows who the Tollense warriors were. At the same time we know, that they have full knowledge since at least 2013:

A pilot study carried out by the applicants confirmed that the preservation of mitochondrial and nuclear DNA is excellent.
http://gepris.dfg.de/gepris/OCTOPUS;?context=projekt&id=236942581&language=en&task=showDetail

I wonder what would happen if e.g. Reich would ask for access to the bones. Fire in a museum?

Samuel Andrews said...

Mix of I1, R1a, R1b in Norse makes it look like they're a mix of multiple Late Neolithic populations.

Davidski said...

They are. Single Grave, Battle-Axe, TRB...

EurDNA said...

Interesting post. yyou're partially right, within the basis of an autosomal statistical excercise - ther's no huge population replacement, I don;t think anybody would suggest otherwise.
However, moving beyond autosomal bean-counting devoid of context, because it there is something of a 30-50% admixture into Sweden after Battle Axe (depending on individual), detectable genetically, with male uniparentals, and most importantanly, the significant social transformations. In fact, it could even be that with the NBA, Scandinavia became IE for the first time

Davidski said...

@EurDNA

In fact, it could even be that with the NBA, Scandinavia became IE for the first time.

Right, so Battle-Axe/Corded Ware wasn't IE?

Lots of problems with that theory, including the data from the East Baltic and South Asia, which point to connections with Battle-Axe and between far flung Indo-European peoples.

As for bean-counting. Try counting how many Germanic beans the people from the Estonian stone-cist graves have. They have none.

And yet the Estonian Bronze Age shows such strong archeological links to Scandinavia that it's sometimes classified as part of the Nordic Bronze Age.

Davidski said...

And Nordic_MN_B isn't Battle Axe.

EurDNA said...

My point was the Nordic BA can be more solidly linked to an I.E. society, whilst BAx is just a LN, tribe society.
I'm not sure what your point about Estonia is, however I suspect the very modest figures you uncovered relate to the use of GAC and even local TRB- which did not exist during the NBA. Use of more proximate modelling might reveal dynamics more clearly

the dude said...

Of the Scandi nations only Denmark has a pretty likely case for beaker intrusion. Sweden has none and Norway is ambiguous as to whether its influence or intrusion. Denmark's beaker is generally thought to mostly be a derivative of Dutch with perhaps some German in the south. So, if beaker was to have any impact on Scandi nations, it would appear to most likely be in Denmark and coming from Holland. Would Dutch beaker intruding into Battle Axe Denmark be detectable on an autosomal level?

Davidski said...

@EurDNA

My point about Estonia was that its Bronze Age population was practically a part of the Nordic Bronze Age, and yet they were obviously native to the East Baltic.

So it's more evidence that the Nordic Bronze Age could spread without any accompanying gene flow.

And it doesn't matter that TRB didn't exist during the NBA. I'm just using it to model the homogenization process within Scandinavia, which, like almost everywhere else, continued for thousands of years after the expansions of CWC into farmer territory.

Wastrel said...

Just a note that I think is important, given the repeated mentions above of "Germanic"...

None of these people were Germanic.

The Proto-Germanic period - in which definably "Germanic" people spoke similar dialects of a common Germanic language - took place between around 500BC and around 400AD (after which it's more plausible to talk of distinct Germanic languages rather than dialects of Common Germanic). These samples are all from centuries, in two of the three cases closer to a millennium, before Germanic.

Recognisably Germanic groups don't enter the record until the 1st century AD, in southern Germany.

At most, the people of northern Denmark and southern Sweden in the Nordic Bronze Age MIGHT be 'PRE-Germanic', in that they might have spoken a language ancestral to Germanic. But it's really important to point out that at this point, there wouldn't have been much specifically 'Germanic' about it at all - it would have been a dialect of a common northwest indo-european that also included pre-celtic and pre-italic dialects, along with many others no longer represented. They would probably have been mutually intelligible, or at least close to it - no more distant than, say, Spanish and Portuguese - and there would probably have been a continuum of transitional dialects linking what we in hindsight call 'pre-celtic', 'pre-italic' and 'pre-germanic'. This 'pre-Germanic' would have lacked the striking features that distinguish later Germanic from other IE languages (like Grimm's Law).

It's plausible that pre-Germanic may have been spoken within the 'Nordic Bronze Age' area, but no clear equation can be made (we don't even know that everyone in the NBA spoke the same dialect). Given how similar everyone IE's language AND genes must have been at that time, it wouldn't have taken much of an elite shift in prestige for pre-Germanic dialects to be spread INTO the area from nearby. Conversely, even if the 'Germanic Migrations' may have begun there, they could easily have picked up the pre-Germanic language en route. It would be very difficult for us to spot this.

Consider how much later 'Celtic' areas look to have continuity with Bell Beaker people. And yet we know linguistically that Celtic must have migrated into the region much later, and archaeologically we can see what we assume to be that migration. Again, we're not talking about clearly linguistically and genetically distinct groups here - we're talking about, at most, cousins, all forming a continuum.

Davidski said...

@Wastrel

My use of the term Germanic above was in the context of claims made in the past that the people buried in stone-cist graves in Estonia were Proto-Germanic-speaking migrants from the Nordic Bronze Age in Scandinavia.

I don't know whether they spoke Proto-Germanic or not. I guess I should have used another term like southern Scandinavian.

Davidski said...

@the dude

Would Dutch beaker intruding into Battle Axe Denmark be detectable on an autosomal level?

At that time Denmark wasn't just home to Battle Axe, but also Single Grave and probably all sorts of post-TRB groups.

But there's practically no difference between the Nordic_LN:RISE98 sample from southern Sweden and the Dutch Beakers. So I don't think the pre-Beaker Danish population was much different from the Dutch Beakers.

However, RISE98 belongs to R1b-U106, while the Dutch Beakers, like all Beakers, are rich in R1b-P312. So I'd say that we're looking at Post-Single Grave genetic affinity, although it's impossible to be sure without Single Grave samples.

Davidski said...

More importantly, who brought Sintashta chariots and horses to the Nordic Bronze Age?

This scene is straight from the MLBA steppes of the southern Urals.

Trundholm sun chariot

FrankN said...

Interesting post, Dave!

Some caveats: The 1700 BC date for the onset of the Nordic BA looks somewhat arbitrary. There was quite some bronze present in Scandinavia already during the LN - mostly halberds imported from Unetice. A massive up-tick of bronze finds, and sustained indication of local production, rather occurs around 1600 BC.
https://www.academia.edu/1909413/Moving_metals_or_indigenous_mining_Provenancing_Scandinavian_Bronze_Age_artefacts_by_lead_isotopes_and_trace_elements

A major cultural shift in S. Scandinavia & N. Germany occured around 2350 BC. In Danish (and Swedish?) archeological nomenclature, this shift marks the transition from the MN Single Grave Culture to the LN Flint Dagger Culture (FDC). The main changes are:

1. Rapid (within a century or less) disappearance of Battle Axes from the archeological record, to be replaced by flint daggers. That phenomenon isn't restricted to S. Scandinavia, but equally observed in N. Germany, the Netherlands, and the Middle Elbe-Saale (MES) region, and is obviously related to BB influence [For that reason, some Danish archeologists now propose to re-name FDC into "Nordic BB" or similar.] Danish Flint Daggers, beautifully-crafted and made from luminiscent, honey-coloured flint, became a major export commodity especially to the Netherlands but also N. England, and seem to have to have made Denmark quite wealthy. With some delay, N. Germans also tapped into to the business and took over a lot of the Dutch/ English market, before ultimately metal (bronze) daggers, and subsequently halberds took domination by at latest 2000 BC.

2. A massive increase of house sizes, to occasionally as large as 400 m2, and even for smaller houses regularly around 150 m2. Essentially, this is the birth of the classic Germanic 2-3 aisled longhouse as well known a/o from the Gothic Wielbark Culture, and prevailing in rural N. Germany (also Netherlands?) until the 1940s [Mine, dating to at least the late 17th century, occupies a moderate 200 m2]. Large longhouses are also characteristic of Unetice, but in this case MES developments seem to post-date the Danish trend.
https://www.academia.edu/19633857/Crises_what_crises_Innovation_Different_approaches_to_climatic_change_around_2200_BCE

The longhouse phenomenon cannot be attributed to CWC. Long believed to haven't posessed permanent houses at all, recently a number of CWC houses have been archeologically identified. However, they tended to be rather small and lightly-built, i.e. lacking the central row(s) of beam posts so characteristic of Unetice and Nordic LNBA longhouses. OTOH, the Schönfelder Culture around Magdeburg - a GAC West offspring hardly affected by CWC - maintained the EN (LBK) longhouse tradition. It's final, "Ammersleben" phase is now dated to ca. 2400-2100 BC, with attestations from the Elbe-Weser triangle via Bohemia down to Lower Austria, S. of the Danube, as such foreshadowing Unetice's reconstitution, and dominance of the Elbe-Danube communication line well known from the EN, but broken during the later MN by the antagonism between Baden and GAC/Bernburg.


Against this background, may I ask you, Dave, to conduct some further analyses on the relation between RISE71 (Danish Single Grave, Zealand) and RISE61 (FDC, Jutland) in order to better understand the quite fundamental change around 2350 BC? More specifically:
(a) Since archeologically, neither TRB nor Danish Single Grave qualify as origin of the BB phenomenon (albeit Danish Single Grave pottery looks quite proto-typical for maritime BBs), are their other, more south-westerly sources that fare well in explaining the difference between the two samples?
(b) Is there indication of a putative Schönfelder Culture influence on FDC? [Schönfelder was a cremating culture, so we will hardly ever get aDNA from them, but GAC Poland plus Esperstedt_MN, maybe also Blätterhöhle, might provide some kind of proxy. Otherwise, Quedlinburg_BB is geographically and chronologically close to the late, Ammersleben phase of Schönfelder.]

Davidski said...

@FrankN

You should read this...

Beaker Longhouses: Livelihood Specialization and Settlement Continuity in North Jutland

Also, how did you work out that RISE71 was a Single Grave sample? I've never seen this individual classified like that. And I'm certain that RISE61 has nothing to do with the FDC. That's the Nordic_MN_N sample that might have something to do with the Single Grave.

In any case, as you can see in the PCA above, RISE61 and RISE71 are very similar.

EurDNA said...

@ Davidski

''My point about Estonia was that its Bronze Age population was practically a part of the Nordic Bronze Age, and yet they were obviously native to the East Baltic.''

Some definitions include Estonia in Nordic BA, sure.
But in fact, it demnstrates 2 wholly different processes.
Estonia Bronze Age is the result of CWC merging with local hunter-gatherers.

That is not the case in Scandinavia proper- the admixing groups were not HGs, but rather a very different sort of people.

''So it's more evidence that the Nordic Bronze Age could spread without any accompanying gene flow. I'm just using it to model the homogenization process within Scandinavia, which, like almost everywhere else, continued for thousands of years after the expansions of CWC into farmer territory.''

This was not necessarily an in situ homogenization process.

I suspect your intention is to demonstrate uninterrupted continuity between CWC & later proto-Germanic speakers. That is not the case even, even with anachronistic modelling.
There is a very good reason why Nordic LN & IA plot where they do [1], and it has nothing to do with CCC people. Rather, it has to do with the movement of groups & elite individuals in privileged positions, who assumed control of a new society in Scandinavia






EurDNA said...

@ Frank

Your point ties in with ''The Dude's'' question.
There were Post-CWC/ BAx/ SGC movements from the continent, via Denmark to parts of Ssweden & Norway.
Indeed, Norway was barely touched by BAx or even TRB, thus the Flint Dagger period you mention is essentially the first arrival of productive economy in the region.
As I have outlined above, I suspect we might be looking at complex processes of mobility and social change, at least a couple of ''waves'' ( not to say these were necessarily discreet or sweeping, but more ''additive''). And these can be easily worked out with more sampling and a bit more effort to look at fine-scale details

Davidski said...

@EurDNA

This was not necessarily an in situ homogenization process.

Well that's what the data suggest very strongly.

In fact, if you look at the full output that I linked to, I don't even need to use TRB to model Nordic_BA as Nordic_LN, or Nordic_LN as Nordic_MN_B.

But TRB improves the fit, probably because of a little more local farmer ancestry and, in the case of Nordic_BA, a slightly more southerly genetic character (2/3 of the samples are from Denmark).

So my models aren't anachronistic. They're actually very sensible, because they take into account the likely uneven levels of farmer ancestry across Scandinavia, which may have even persisted into the Iron Age, like they did in Poland.

This is what you need to accept first and account for it before postulating about any significant migrations of elites or otherwise.

EurDNA said...

For now, I am happy with my working mode, based on the PCA, corroborating Y-hg data, and archaeology.
But I agree it's a work in progress..

Davidski said...

@EurDNA

Looks like a very likely scenario considering that Nordic_BA is basically indistinguishable from Nordic_LN, and Nordic_LN from Nordic_MN_B.

Look at the stats. TRB isn't even necessary.

https://drive.google.com/open?id=1V6MglIP4sOojqryKJWvtMZjeYQ-M81GI

https://drive.google.com/open?id=1QEy6NrEJ0xszjPZSJOI1-YD2buHWOwv4

And what is the sanity check provided from Y-hg exactly? How do you know that there wasn't any R1b-U106 or I1 in Scandinavian Single Grave? Where's the data showing this?

Battle Axe is from the East Baltic, so no wonder its R1a, but we don't even know for sure that there wasn't any R1b-U106 in Battle Axe.

Leucuuo said...

John Thomas:
"Everyone is Polish. Only they don't know it : )"

Ha. Before I got my results from Living DNA several weeks ago, I wouldn't have taken that comment seriously. I chose LDNA because my ancestry is, anecdotally, British and Irish. My y-dna hg is R-DF13 – so no surprise there. And according to the Eurogenes K36 calculator at GEDmatch, my aDNA is overwhelmingly "North Sea" and "Atlantic" – followed by the "Iberian", "French", "Italian" and "Fennoscandian" components often found in modern Brit/Irish pops.

However ... my mtDNA is U2e, which from my reading may have originated with the Kunda/Narva neolithic. This is paralleled by other Eurogenes K36 components: 9% "Central Euro", 6% "Eastern Euro" and 4% "East Central Euro". Perhaps some of this can be explained by medieval migration/slave trading into Scandinavia. But I'm thinking it's more likely (in light of recent publications discussed at this forum) that these represent residual EHG/CWC components in subsequent Beaker populations that headed west until they got to Ireland?

FrankN said...

Dave: Thx for your link - I'll certainly read it diligently, as I have a strong personal interest in figuring out more about the origin of Germanic (Saxon) long houses. As a first note: This is one example of the trend among Danish archeologists to re-define the Flint Dagger Culture as Scandinavian expression of the BB phenomenon.

Otherwise, I have messed up samples and their IDs - sorry! Of course, RISE71, Nordic LN, 2196-2023 BC, is Flint Dagger Culture, and RISE61, Nordic MN B, 2851-2492 BC, is Single Grave (although, looking at Kroonen/Iversen 2017, we might still be dealing with late TRB here).
https://static-curis.ku.dk/portal/files/184141394/AJA121_04_Iversen.pdf

"In any case, as you can see in the PCA above, RISE61 and RISE71 are very similar."
If my eyes don't deceive me (which may well be the case, I am getting into the age when eye-glasses are needed), RISE61 and RISE71 plot in quite different positions in your PCA. RISE71 (Flint Dagger Culture) is displaced towards where in your PCA the labels "Belgian" and "French" appear. Of course, a PCA only captures two dimensions, the aDNA reality is multidimensional - hence your G25, for which I am extremely grateful.
As such, to not just base suggestions on two-dimensional analyses and my dwindling eye-power, let me re-iterate the request to analyse the relation between RISE61 and RISE71 by formal methods…

Otherwise, I sense some feeling of antagonism in your reply. I am certainly not always according with your opinion (the "Botai-like Central Asians" are an issue that I will address in due time in an own post over at aDNAera), but when it comes to LNBA Scandinavia, and the implied topic of Pre-Proto-Germanic, I think we essentially share the same line of thought. I have expressed my respective ideas clearly as comments here:
https://bellbeakerblogger.blogspot.com/2018/07/schnursprecher-glockensprecher.html

My interpretation of your analyses is a confirmation of these ideas, i.e. linking the emergence of pre-Germanic to Single Grave - TRB interaction, since there is little evidence of post-EBA demic change in S. Scandinavia. The open question to me is the effect of the 2350 BC transition on this process - a question on which I hope for your assistance. But that is more a technical than a fundamental question - essentially I am with Kroonen/ Iversen 2017, which implies CWC/ Single Grave as the agent that brought IE to S. Scandinavia.

EurDNA said...

@ Davidski

''Looks like a very likely scenario considering that Nordic_BA is basically indistinguishable from Nordic_LN, and Nordic_LN from Nordic_MN_B.''

I agree, that globally theyre all very similar.
However, I personally am also interested in specifics, e.g.

Nordic_LN:RISE98
Beaker_The_Netherlands 49.9%
SWE_Battle-Axe 26.4%
SWE_TRB 11.7%

Nordic_IA:RISE174
Beaker_The_Netherlands 44.7%
SWE_Battle-Axe 23.3%
Unetice_POL 19.5%
Baltic_HG:Spiginas4 6.5%

And, given that neither I1 nor U106 nor I2a2 (some of the dominant 'proto-Germanic'markers) are found in BAx or TRB- there must be somethign more going on here.

I admit that I am not quite clear about the SGC- BB transition, and that they might all be of ultimately similar origin, but that's a side issue

Ric Hern said...

So basically dialect levelling could have taken place from +- 2700 to 1700 BCE giving Germanic its distinctive characteristics....?

Samuel Andrews said...

In total distance of G25 positions in ancient Scandinavians....

RISE98, R1b U106+, 2100bc, Sweden. Has clear Scandinavian affinities. Almost all of his closest modern matches are Swedes.
RISE61, Denmark, 2600bc. Also has Scandinavian affinities. But not as obvious. He's closet to RISE98.

Martin said...

I don't think anybody asked, but just let me contribute my Nordic Prehistorian 5 cents:

In Scandyland, "LN" means 2350-1700 cal BC. And "BA" means 1700-520 cal BC.

Thanks!

ambron said...

David, would you reveal Y-DNA from Tollense if they tried to make it secret?
Arza's forecasts do not comfort us.

Davidski said...

@ambron

David, would you reveal Y-DNA from Tollense if they tried to make it secret? Arza's forecasts do not comfort us.

The paper is definitely coming with all the details. I don't know why it's taking so long, but possibly because the authors want to be very thorough.

Davidski said...

@Martin

In Scandyland, "LN" means 2350-1700 cal BC. And "BA" means 1700-520 cal BC.

Yes, and that Corded Ware-related sample (RISE61) with a tonne of steppe ancestry is actually classified as MN.

I initially thought it was a TRB sample because of that.


epoch said...

@EurDNA

"And, given that neither I1 nor U106 nor I2a2 (some of the dominant 'proto-Germanic'markers) are found in BAx or TRB- there must be somethign more going on here."

According to the Yfull page of I1 its date of origin is 27.500 years ago but its most recent common ancestor is 4600 years ago. That means it was "in hiding" for a really long time, only to expand somewhere in the 3rd millenium before Christ. That fits the expansion of BB, or even CWC. IIRC its highest diversity is roughly the North Sea coast.

To me it seems there is a real possibility that I1 was a minority haplogroup among TRB which profited from the BB expansion, in which case we might never find a sample of it. The oldest I1 has been found in LBK from Hungary, so it did have a presence in farmers.

Davidski said...

@EurDNA, Frank and Samuel Andrews

Keep in mind that some of those ancient Nordic samples in my PCA above are heavily damaged, and even show clear signals of pseudo African ancestry because of that. So it's important not to focus too closely on where they plot in fine scale analyses, especially those based on two dimensions or PCs.

But despite their damage and low quality, it's clear that they are very similar to each other and to the Dutch Beakers.

They all show very strong signals of Germanic-specific genetic drift, just like the Dutch Beakers do. The Global25 picks this up well.

My explanation for this phenomenon is that, apart from the Battle Axe samples, they're all in large part derived from Single Grave people, and that the Single Grave people are the largest contributor to the genetic structure of Germanic populations, especially northern Germanic populations.

But hey, I could be wrong. It's happened before. If so, there must be another explanation, but I don't have a clue what it might be at this stage.

weure said...

Rise61:

"Judging from the 13C collagen values, this is likely to only affect
one case, namely sample RISE61, from Kyndeløse in Denmark. It is a young man with a highmarine signal, and his dating should likely be reduced by a couple of hundred years. Freshwater reservoir effects, as can be deduced from combined 13C and 15N values, are also of marginal importance in these data.'

"RISE61 Denmark Scandinavia baSca"

"4071 BP"

Source: https://media.nature.com/original/nature-assets/nature/journal/v522/n7555/extref/nature14507-s1.pdf

ambron said...

David, clear! They must be very thorough in this situation.

Davidski said...

@EurDNA

And, given that neither I1 nor U106 nor I2a2 (some of the dominant 'proto-Germanic'markers) are found in BAx or TRB- there must be somethign more going on here.

I can think of two things: post-Battle Axe/TRB local founder effects and lack of samples.

How many Battle Axe and TRB Y-chromosome samples do we have available at this stage?

Can you list them for me?

weure said...

correction

Blogger weure said...
@Frank I plot close to Rise71 (not 61). I guess this is a continuing episode of your TRB-Tiefstich expansion. Along the North Sea and somewhat more inland, on sandy soils, we see this TRB/SGC integration proces. Those longhouses were indeed typical for the NE Dutch sandy soils too....even until 1900 indeed. The clay districts like Dithmarschen is of course another story.
And let’s not forget that Germanic is a Roman or Classic world invention not for anthropological reasons. Genetic and language are not always 1:1.
But one thing seems very remarkable in the Ingevaenic area there seems to be a genetic continuity from LN-EBA until now....

Angantyr said...

TRB (North) and Battle Axe Y-DNA:

Gok4: I2a1b1 (according to Genetiker)
Ansarve 8: I2a1b1a1
Ansarve 14: I2a1b1a1
Ansarve 17: I2a1b1a1
Ansarve 16: I2a1b
--
RISE94: R1a1(a1b - Genetiker)
Olsund: R1a1a1b

And then there's RISE61 without a clear archaeological culture assignment. He grave was a secondary burial in a (TRB) passage grave on Zealand which never became either Single Grave or Battle Axe territory, but he's clearly not a late TRB survivor either. He's R1a1a1(b1a3b1 - Genetiker) anyway.

weure said...

@Angantyr that's a pattern we see in the North Dutch area (Drenthe) too...TRB dolmens are re-used in later periods (BA).

Davidski said...

@Angantyr

Yep, so currently we have a handful of Battle Axe and TRB samples from a couple of sites, and most of these samples are from Gotland.

EurDNA said...

In addition,
TRB - Baalberg I2a1b
TRB-Salzmunde I2a1a
But, TRb - Poland : C1a (probably a ''farmer'' lineage here)

@ Davidski
What we've learned so far is that these patterns generally tend to hold.
But your first point is right, there was a post-BAx expansion, which has been my contention all along. And this was not some random ''bounce back'', but one tied to rather specific , deliberate phenomena, as ive already outlined

@ Epoch
''To me it seems there is a real possibility that I1 was a minority haplogroup among TRB which profited from the BB expansion, in which case we might never find a sample of it.''

That would be a rather odd outcome. From what we've seen, BB was an exclusive L51 phenomenon, and non-BB doen't seem to have exactly ''profited'' from them lol. WE have to remember, or perhaps learn anew, that the history of Europe did not end with CWC - BB horizon, even if some academics tend to think so.

''The oldest I1 has been found in LBK from Hungary, so it did have a presence in farmers.''
That might make more sense http://tinypic.com/view.php?pic=eaobpz&s=9#.XOJzNcgzbD4
, but still hypothetical I admit


Davidski said...

@EurDNA

The high frequencies of U106 and I1 in Scandinavia are best explained by extreme local founder effects during the Bronze Age.

Not migrations of U106 and I1 rich populations into Scandinavia from, say, the Carpathian Basin.

It's unlikely that anyone will find the relevant, ancestral I1 in the ancient DNA before the founder effect. That is, dating to the TRB, Battle Axe or BB periods.

Samuel Andrews said...

A common Single grave origin for R1b P312, R1b U106, I1 clans could explain the northwest European-specific drift found in DNA tests. What else links British/Irish Celts & Scandinavians?

EurDNA said...

@ Davidski

Well they were founder effects, but not random ones, anymore than we can fob off Z645 being a random founder effect.
They don't have to be from Hungary, but the (admittedly paultry) one lead we do have for I1 points there, as does a host of other evidence. U106 could be much closer to Sweden
Its one thing you are unaware of these matters, but its another to arrogantly dismiss them when you have no authority to do so.


''It's unlikely that anyone will find the relevant,''

LOL. A rather bold prediction.

Davidski said...

@EurDNA

Right, but you apparently have the authority to dismiss a theory offered by linguists based on linguistics data that Proto-Germanic came from Corded Ware.

Maybe it did or it didn't, but you've shown nothing here to even put a dent it.

At least I can back up my claim that there was strong, and possibly almost complete, genetic continuity in southern Scandinavia from the Nordic_MN_B to the Nordic_BA. Admittedly, there's nothing special about that, since it's a very simple thing to do.

By the way, you need go and have a closer look at the U106 and I1 SNPs. They show signals of extreme founder effects in Scandinavia dating to the Bronze Age and look like they spread to the rest of Europe from there. And this is rather common knowledge too.

Carpathian Basin? LOL

Samuel Andrews said...

@Davidski, single samples are pretty weak argument for continuation though.

Davidski said...

@Samuel Andrews

Single samples are pretty weak argument for continuation though.

Why so?

I'd say the fact that these sample sets are made up of one to three random individuals from different sites in Denmark and Sweden means that we're looking at a million to one sort of thing that the continuity shown from Nordic_MN_B to Nordic_LN to Nordic_BA is a fluke.

Are you saying that these random samples from sites separated by hundreds of years and miles just happen to be closely related to one another by pure chance?

Samuel Andrews said...

Similarity across time for it is good evidence. But If, let's say a population from Netherlands moved into Scandinavia in 1500bc, it'd be hard to tell because the newcomers would be so similar to people already living there. So, obviously new samples are needed to confirm this.

The theory of Single Grave, Battle Axe, TRB origins for Scandinavians & continuation since Late Neolithic also make sense because it seems a lot of Europe's gene pool got established in late Neolithic/Bronze age.

Vinitharya said...

@John Thomas "Everyone is Polish"
I don't know about that; I have a Polish last name and a R1a-M458 Y-chromosome, but a glance at my autosomal data would quickly disabuse you of that notion (my father and his two brothers took their stepfather's last name when they moved across the country from California to Florida, to start a new life, I guess, and he was first generation, too, my Nana said that her in-laws were very Polish peasant, especially seeing as their last name came from a dialectical term for a type of buckwheat porridge). It could be that haplogroup I1 came to Scandinavia with R1a-Z284, as it has really thrived too much to be a despised conquered peasant lineage. In Germany the farmer lines were decimated and the HG lines are a minority, if well-represented. Germany is now dominated by R1b Bell Beakers, the rather latecoming R1a Wendish Warriors, and also the I1 coming from Scandinavia with the Proto-Germans.

weure said...

I guess the so called Elp culture is no indcident, this follows exactly the SGC/TRB hotspots around the North Sea, from Denmark, NW Germany to the Northern Netherlands.

https://en.wikipedia.org/wiki/Elp_culture

No coincidence that I, from North Dutch stock, plot right in the middle of this LB/BA Nordics samples (close to Rise71).

Th only culture of that period that came unto the scenethat could have caused immigration from elswhere is the Sogel Wohlde culture, they brought the Tumulus to the Elp area, in fact constitueted the Elp culture. But I dont know if this was acculturalization or if this has had genetic implications.

In the late Bronze Age the Valsomagle culture expanded from East Denmark/Scania towards the Elbe. They were in fact pre-Jastorf

Simon_W said...

Modelling the Elbe Germanic Baiuvarii as a mix of Nordic_BA and Hallstatt:

[1] "distance%=3.0476"

DEU_Medieval

Nordic_BA,65.4
CZE_Hallstatt_Bylany:DA111,34.6

Arza said...

@ Davidski

Tollense guys are knocking at the door:

Nordic_BA
Nordic_LN 48%
SWE_Battle-Axe 30.8%
HUN_LBA:I1504 21.2%
Distance 2.2823%

Nordic_BA
Nordic_LN 83%
HUN_LBA:I1504 17%
Distance 2.7543%

How does this look in qpAdm?

Davidski said...

@Arza

Not really any better than the models I posted. As per my blog entry...

Of course, if I mixed and matched reference populations from across prehistoric Europe, I could probably come up with some spectacular statistical fits...

Simon_W said...

To me using HUN_LBA:I1504 doesn't look like a sensible way of taking potential influence of Tollense guys into account. Because I1504 doesn't have that strong WHG shift that many of the Tollense guys have. I1502 is much closer to them in the West Eurasian PCA. And one of the Vatya samples, but I forgot which one.

But speaking of the Tollense guys, in Dave's North European PCA some of them overlap with Germanic people and some with Slavs. I'd say the most natural interpretation is that the Germanic-like ones came from the Nordic Bronze Age and the Slavic-like ones from the Lusatian culture. Which would entail the interesting conclusion that the Lusatian culture had accumulated a lot of Slavic-specific drift. But then again, in the West Eurasian PCA, there's this peculiar WHG shift that many of the Tollense guys have. Some of them overlap with Germans and Poles, but many don't overlap with any modern population in the West Eurasian PCA.

André de Vasconcelos said...

You are probably right, but as of now I'm hoping you're not, just so we can see you eating your own hat online :)

Wastrel said...

Davidski: just to clarify, my comment about Germanic was directed less at you and more at some others who seemed to be getting over-enthusiastic. It's reasonable, of course, to use 'Germanic' as a shorthand in some contexts. But I think some people (understandably) fall into the trap of projecting modern groupings backward, as though they were discrete and coherent blocks, even into time-periods long before they have any validity.

weure said...

@Wastrel absolutely right Germanic is Roman/ Iron Age label, not before. Nevertheless there seems on the Northern Plain and South Scandinavian room a kind of genetic continuity since LN/EBA. Davidski has shown that very clear IMO.

FrankN said...

EurDNA: "They don't have to be from Hungary, but the (admittedly paultry) one lead we do have for I1 points there, as does a host of other evidence."

But note that Blätterhöhle, IIRC also Motala prefer Hungary_HG over Loschbaur in G25 models. The Laacher See eruption turned everything between Middle Rhine and Middle Elbe into barren, uninhabitable land for quite some time. There is, e.g., not a single early Mesolithic site known from otherwise archeologically well studied Thuringia.

As such, we probably need to envisage at least three, somewhat genetically differentiated streams of post-Younger Dryas repopulation of the more northerly parts of WC Europe:

1. Franco-Cantabrian - represented by La Brana etc.,

2. Thyrrhenian: Bichon, Loschbaur etc.

3. Adriatic Sea/ Carpathian Basin: East of the Rhine and the Laacher See fallout zone, including, as archeologically indicated, Ahrensburgian and derived Scandinavian Mesolithic pops. Villabruna R1b obviously belongs here, and I1 might as well.

[I am uncertain about the status of IronGates_HG in this respect. It seems to incorporate some minor Barcin and CHG admix. This admix may date already to the UP - then IronGates constitutes another repopulation source in its own right, especially as concerns Mesolithic Ukraine. However, the Barcin and CHG elements in IronGates may also date to the Mesolithic, in which case the pre-admix population would probably count as part of the Adriatic/ Carpatian basin repopulation stream listed as 3) above.]

In any case, there is IMO nothing speaking again "Carpathian HG" presence, including yDNA R1b and I1, already in Ahresburgians / Maglemose/ Kongemose/ Ertebölle, none of which has yet been successfully tested for aDNA [M. Allentoft apparently tried it, but Danish soils don't seem to conserve aDNA well].

Grey said...

i think "people not pots" is the general case for most of history as if a population developed an advantage over their neighbors they'd likely expand into their neighbor's territory but one big exception to this would be situations where an ecological boundary which makes such an expansion physically impossible.

one example of this kind of scenario might be farmers (or herders) not being physically able to expand into wetland regions so i'm wondering if the general case in this specific kind of situation was pots not people i.e. something like Ertobolle with the herders/farmers expanding up to the edge of the wetlands and a hybrid population/culture developing among the wetlands population in the contact zone.

if correct although the farmers couldn't physically utilize wetlands (until drained centuries later) and the wetlands population couldn't fully adopt farming (until drained centuries later) there are often lots of small islands in wetlands where sheep, pigs, cows etc could be grazed so it wouldn't be surprising if the wetlands foragers adopted that aspect of farming.

if correct then the thing about this is adding domesticated animals to their food package would have given an advantage to that part of the wetlands population in the contact zone over that part of the wetlands population further away from the contact zone leading to the possibility of what you might call an "internal invasion" from the edge to the interior leading to the homogenization suggested in the OP.

a more modern analogy might be two genetically similar but culturally distinct tribes one on the coast and one further inland where the coastal tribe gets guns from foreign traders and uses them to first conquer and later assimilate with the inland tribe.

if correct an "internal invasion" from contact zone to the interior might have happened twice: fist time sparked by the farmer expansion (mostly south to north?) and a second one from the herders (mostly east to west?)

#

various randomly connect thoughts

Vanir, Wends, (multiple) Veneti, I1 etc = river deltas / wetlands?

Black Sea and Baltic wetland populations connected by the big rivers with little impact on the non wetlands population in between?

Davidski said...

@FrankN

You're going off on some weird tangent that has nothing to do with the blog post.

The high frequencies of U106 and I1 in Scandinavia today are the result of founder effects and population growth during the metal ages and the Migration Period.

FrankN said...

Dave: "The high frequencies of U106 and I1 in Scandinavia today are the result of founder effects and population growth during the metal ages and the Migration Period."
No fundamental dissense here - albeit I think the population growth rather occured before the Migration Period, and motivated respective migrations. In fact, the Cimbri leaving Himmerland (N. Jutland) in the late 2nd cBC, Goths leaving Götaland (around Gothenburg) around the same time [to the extent these etymological connections are correct], Ariovist's Suebians finding their way from the Lower Oder (Suebia in antique sources) to SE Gaul during the early 1st cBC, and Germanic Chatti (->Hesse) contemporarily replacing Celtic Ubii in N. Hesse, points IMO to substantial Germanic population growth followed by outmigration around 150-70 BC. Another growth phase during the 3rd cAD is archeologically well attested from sites like Feddersen Wierde near Bremerhaven.
[A main factor seems to have been the switch to rye, better climate-adapted than wheat, as main crop. The cultivation history of rye is yet poorly understood. It seems to have originated in E. Anatolia, from where it spread via the Caucasus into the Steppe and E. Europe, but the timing of that dissemination is yet unclear. Might have been the BA, but could also have been as late as the early IA (Skyths, proto-Slavs). In any case, rye wasn't yet part of the Yamnaya/CWC/SGC "agricultural package" that instead focused on wool sheep and non-light sensitive barley, both of ultimately Iranian origin.]

My point was directed towards EurDNA and intended to indicate that genetic connections between the Carpathian Basin and N. Europe may predate the Neolithic.

EurDNA said...

@ FrankN

This is very interesting but somewhat tangential, so I cant dwell on them here to deeply.

''As such, we probably need to envisage at least three, somewhat genetically differentiated streams of post-Younger Dryas repopulation of the more northerly parts of WC Europe:

1. Franco-Cantabrian - represented by La Brana etc.,

2. Thyrrhenian: Bichon, Loschbaur etc.

3. Adriatic Sea/ Carpathian Basin:''


It might be that much of what we thought we knew is wrong. The F-C refuge is not so prominent as believed, there is no such thing as a Tyrrhenian refuge, but indeed, the Villabruna - WHG cluster is probably ''modal'' around the Adriatic (Italy - Balkans). This is because it expanded from there with the Late Glacial Epi-Gravettian -> Azilianization process.
We have aDNA from Ahrensburgian descendants, buy way of the Nordic Tanged-Pointed industries - some very early Mesolithic from Norway & Sweden - they're all I2a apart from one or 2 I2c.
THere might have been R1 in Moglemose, but frankly I think this is a very slim shot - as R1b is mostly related to southern & eastern Epigravettian groups.


''My point was directed towards EurDNA and intended to indicate that genetic connections between the Carpathian Basin and N. Europe may predate the Neolithic.''

Frank, that's a pretty silly statement, how do the two exclude each other ?
I think I can tell the difference between Mesolithic and a Bronze Age expansion.

EurDNA said...

@ Davidski

''Right, but you apparently have the authority to dismiss a theory offered by linguists based on linguistics data that Proto-Germanic came from Corded Ware.''

Davidski, linguists generally agree that proto-Germanic is an Iron Age language. In fact, the sound shifts which characterise Germanic were still developing during the Roman Era. The expansion of Germanic peoples occurred in the wake of the La Tene collapse
Nothing you have offered here contradicts that, and in fact, you seem blissfully unaware of the fundamentals

''At least I can back up my claim that there was strong, and possibly almost complete, genetic continuity in southern Scandinavia from the Nordic_MN_B to the Nordic_BA''

And I don't deny that. But I am pointing out that is too broad a, zoomed-out perspective,
Everybody in northern Europe will is some shape or form descend in large part from CWC & post-CWC, but that's not really telling us much specifics is it ? Qute honestly, Id treat the Kroonen/ Kristiansen model as a very preliminary one, putting it politely.

''By the way, you need go and have a closer look at the U106 and I1 SNPs. They show signals of extreme founder effects in Scandinavia dating to the Bronze Age and look like they spread to the rest of Europe from there. And this is rather common knowledge too.''

Yes U106 and I1 appear according to modern distribution to have expanded from Scandinavia. But unless you're suggestng that U106 is from Mesolithic Scandinavia, that doesn't negate the probability that it first of all had to arrive from somewhere else, and then locally expand, and then further afield in later prehistory

And whilst we''re on the topic of common knowledge, I'm sure you're aware that the frequency of R1a-L664/ Z282 in Germanic speakers is not very prevalent: 6% in Dutch, 11% in southern Sweden, ~0% in Germans.
Whoops.

Davidski said...

@EurDNA

Nothing you have offered here contradicts that, and in fact, you seem blissfully unaware of the fundamentals.

I'm not trying to contradict or dismiss any mainstream linguistics theories. You are.

The theory that Proto-Germanic was derived from the language of the Corded Ware people has been around for a long time and there's at least one recent peer-reviewed linguistics paper published on the topic.

The fact that ancient DNA shows very strong genetic continuity in Scandinavia from the Corded Ware period to the NBA obviously corroborates this linguistics-based theory.

It doesn't prove it, because genetics can't prove or debunk linguistics theories, it can either corroborate or contradict them.

You seem to think that you can debunk a linguistics theory with your interpretation of the origins and histories of Scandinavian-specific Y-haplogroups. I'm telling you that you can't.

First of all, your interpretation of the Y-haplogroup data is your own, and doesn't necessarily reflect reality, and secondly there is no rule that says shifts in Y-haplogroups should correlate with shifts in languages.

Take a look at the ancient Iberian data. It suggests that Iberian and Vasconic speakers had frequencies of R1b-L51 of up to 100%. How did this happen? No one knows for sure. You'll soon see basically the same thing in ancient data from Italy, with Etruscan speakers being identical to early Italic speakers.

You might think that you've come up with a brilliant theory to explain how Germanic formed in Northern Europe based on your own interpretations of all sorts of data, but honestly I'm not impressed with anything you've argued here.

Al Bundy said...

@Davidski Thanks for all your hard work, when is the Italy preprint or paper coming out?

Davidski said...

@Al Bundy

I don't know, but maybe soon from what I heard.

The results are interesting but not overly surprising, with early Italic and Etruscan speakers showing quite a bit of steppe influence. Romans, and especially late Romans, show much less steppe ancestry, and there are some outliers among them that plot in the Middle East.

Andrzejewski said...

@Davidski "The theory that Proto-Germanic was derived from the language of the Corded Ware people has been around for a long time and there's at least one recent peer-reviewed linguistics paper published on the topic."

Actually, I have read papers claiming that the EARLIEST layer of the Proto-Germanic languages was derived from a Satem-type CWC language ancestral to Proto-Slavo-Baltic and their ilk. Centuries later a Centum-style language closer to Italo-Celtic made a much larger impact of the emergence and evolution of the Proto-Germanic ethnogenesis. The NBA was allegedly the time when the CWC/SGC and the Bell Beaker fused to create the Germanic branch. Therefore Germanic stands out as a unique and unusual branch of Indo-European with its own distinct vocabulary, morphology and "Grim Laws" of phonology. Many people also attribute these phenomena to the strong underlying impacts of "farmer" and "forager" substrate languages like those of LBK, TRB or Ertebolle but so far I'm not very convinced.

Jim said...

"Since linguistic evidence suggests a heavy Italo-Celtic like influence in Proto-Germanic, this raises the question of how this influence occurred."

Proto-Germanic is in the centuries leading up to the common era, much later than this. And in any case the Celtic influence in Germanic languages is in West Germanic, not North Germanic (Norse) or Eastern Germanic (Gothic).

Italo-Celtic is at best a Sprachbund, not a genetic node.

FrankN said...

EurDNA: "linguists generally agree that proto-Germanic is an Iron Age language. In fact, the sound shifts which characterise Germanic were still developing during the Roman Era.

I have had a similar kind of discussion last year with BBB, see
https://bellbeakerblogger.blogspot.com/2018/07/schnursprecher-glockensprecher.html

Of course, the proto-Germanic sound shifts (Grimm's law) occured rather late, possibly, as indicated by the C(h)atti->Hassi(->Hesse) shift, until after the 2nd cAD. In fact, the K->(C)h sound shift was still at work during the Medieval, compare "Karl der Grosse" (Hung. Karoly and Slavic analogies betray the Old High German "k") vs. "Charlemagne", and can till to date be observed in Schwizerdütsch (Swiss German).

But the process of Germanic differentiating from (late) PIE started long before Grimm's Laws took effect. Pre-proto-Germanic, alternatively also called Germanic Parent Language (GPL), a/o fundamentally reworked the IE tense system. Essentially, it turned the PIE aorist into a general marker of past tense (Latin, in contrast, merged the aorist with the perfect, Bulgarian has maintained it to date), and developed from it the (in)famous set of "strong" or "irregular" verbs (sing, sang sung, etc.).
Another feature is the general simplification of the PIE tense system. Gothic, as modern English, e.g., had lost the future, replacing it with either temporal adverbs ("tomorrow I go") or auxiliary verb constructions similar to English "will" / "shall". There is also no indication of the presence of plusquamperfect forms in any attested Germanic language. Proto-Germanic had lost the conjunctive, which it merged with and replaced by the optative. Also, there is very little, if at all, evidence of an early Germanic passive other than constructions with auxiliary verbs as "to be" ([Old] English) or "w(u)erden" (to become -German, Gothic).

Essentially, these features setting apart Germanic from PIE must have originated long before the proto-Germanic sound shift. How long before, is difficult to tell. Might have been the Nordic BA, might also have been the Nordic MN or LN. As such, the possibillity of Danish Single Grave adopting TRB language, and only (re-)switching to IE under Unetice influence can't be excluded, albeit it is certainly not the most parsimonous scenario.

On thing, however, is clear: Whatever language merged with late PIE to evolve, as a hybrid, into (pre-)proto-Germanic - it can't have been (Ibero-)Vasconic. The non-IE substrate in Germanic is full of roots that entail consonant clusters, while Basque phonotactics prohibit intra-syllabic consonant clusters.

Samuel Andrews said...

Granted, Germanic became fully formed in Iron age, but it had been developing in Scandinavia for two thousand years.

Proto-Germanic's ancestor in 1000bc was a very similar language to later Germanic in 100bc. Maybe there was a kind of para-Germanic sphere in Scandinavia in the Bronze age, many related languages related to later Germanic. Later one language replaced the rest.

So, it isn't completely inappropriate to refer to Late Neolithic/Bronze age Scandinavia as "proto-proto-Germanic."

FrankN said...

@Jim (and others):
The issue of Celtic influence on Germanic is easily explained. Germanics and Celts co-habitatited at least from Central Germany (Hesse) through Bohemia (Celtic Boii taken over by Germanic Markomanni), Pannonia (Germanic Gepids), and SE Poland (archeologically attested Celtic settlement chambers around Cracov well into the 1st cAD). Goths most likely encountered Celts in SE Europe (Gothic atta "father" may either reflect proto-Celtic loss of "p", or encounter with proto-Turks, the former is IMO the more likely scenario). Last but not least, Scotch-Irish monks were instrumental in missionising/ educating South Germany, and in fact the first to record Old High German.

As to Italic - strong Latin influence on High German, and to a lesser extent also English, is obvious and shouldn't require further elaboration. Less well known are the Italic connections of the Pommeranian Face Urn Culture, and the related House Urn Culture around the Harz, both dating to around the 6th cBC. There is quite some indication of Italic substrate in Czech, e.g. the Italic g->h sound shift also known from that language, or (West) Slavic *vrh "mountain" that via Venetic bh->v and gh->h sound shifts can be connected to PIE *bhergh "elevation" (Celtic *brig(a), German Berg). An Italic-derived German toponym might be Jena, from PIE genu "knee, (river) bend", c.f. Genua. The most compelling evidence IMO for a Venetic origin of the Pommeranian Face Urn Culture is the mythological city of Vineta, now commonly identified with Wolin on the mouth of the Oder.
https://en.wikipedia.org/wiki/Vineta

Ric Hern said...

It is interesting how some people want to prove that Indo-European families have nothing to do with each other at a Proto Level for some kind of personal and or collective reason...but at the end of the day the fact that PIE existed and dialects and Genes crossing linguistic borders exist doesn't correlate with this Exclusivity Theory on Proto Level. Yes dialect levelling events most probably connected to shared Cultural practices occurred but it does not mean that there were a total break of all linguistic similarities on a dialect level and from the Overarching PIE similarities. If Proto-Indo-Europeans spread as a group into Central Europe from the East, how can there be no connection of their daughter Languages to each other on a Proto Level?

Ric Hern said...

Hypothetically It is basically like saying PIE were spoken until 2000 BCE in this area and 500 BCE in that area while the Genes associated already arrived 3000 BCE in both areas....

Andrzejewski said...

Give us an example of “non-IE origin”; IMO pre-proto-Germanic if you will was an amalgam of an earlier CWC Satem substrate with a much later and much more profound Italo-Celtic like BB Centum influence.

jan.t.andersson said...

Let's skip this "Germanic" thing. The term was invented by the Romans and they weren't very clear on what they included in it so why would we continue using it. The so-called Germanic culture and language evolved in Scandinavia as a direct continuation of the Nordic bronze age. Sound shift didn't create Germans, they were Scandinavians both before and after some influencer among them developed a speech impediment. So why not just call it Scandinavian.

So we Scandinavians come to rule much of the world.

EastPole said...

@Davidski

“The theory that Proto-Germanic was derived from the language of the Corded Ware people has been around for a long “

Yes, this used to be the mainstream theory. For example Frederik Kortlandt in “THE SPREAD OF THE INDO-EUROPEANS”wrote:

“The Indo-Europeans who remained after the migrations became speakers of Balto-Slavic. If the speakers of the other satem languages can be assigned to the Yamnaya horizon and the western Indo-Europeans to the Corded Ware horizon, it is attractive to assign the ancestors of the Balts and the Slavs to the Middle Dnieper culture.”

But that theory of western linguists has been disproved by genetics now. Satem languages like Balto-Slavic and Indo-Iranian are derived from Corded Ware, not from Yamnaya. Works of Mittnik et al. 2018, Narasimhan et al. 2018, Damgaard et al. 2018 and others support it now strongly.

Corded Ware was Indo-Slavic, I have no doubt about it.

The origin of Italo-Celto-Germanic group is linked to Yamnaya now. But it still needs more evidence.

Davidski said...

@jan.t.andersson

So we Scandinavians come to rule much of the world.

How much of the world do you rule exactly?

Bob Floy said...

They rule a few blocks in downtown Stockholm.

weure said...

@ Jan Andersson, yes why not replace Deutschland by Scandinavia in the German national anthem (the nowadays not mentioned phrase...uber alles in der Welt). Cut that kind of crap.. In this kind of research this is about the worst association we could use IMO :(

weure said...

Frank ‘Another growth phase during the 3rd cAD is archeologically well attested from sites like Feddersen Wierde near Bremerhaven.’

That is intriguing because during the Roman times the people from this area (‘the Çhauken’) later called the Anglo-Saxons (give it a name) were expanding to Groningen and Nord-Drenthe area. And in migration time they expanded also to nowadays Friesland especially Westergo, Westergo was in the fourth century abandoned, but was in the North Sea area a strategic place.

See:
http://jalc.nl/cgi/t/text/get-pdfd43e.pdf?c=jalc;idno=0402a03

This give a mixed kind of population in North Dutch. In parts of Friesland much Anglo-Saxon influx,, in Groningen mixed (A_S influx and BA substrate) and in Drenthe the highest old Elp or BA substrate!

The question is if the genetic profile of the incoming population during the Migration Period was very different from the indigenous LN/EBA, Elp population. Most probably not.

To exemplify this the plot of my family all stocky North Dutch. My mother (EBG) has only ancestry from the TRB hotspot (and also Single Grave) of the Netherlands called Hondsrug/Drenthe. Mom and I are twins of Rise71 and Rise97.Fully in the Nordic LN-BA pattern. As could be expected from the Drenthe area... My father (JB) is a more Anglo-Saxon influx and indigenous combination ( mix from especially Groningen, but also some Friesland and Drenthe) is slightly more Western.

https://www.mupload.nl/img/anb5rzt.png

However, Germanic, or pre of proto or whatsoever Germanic, or not, we see all around the North Sea area a continuous genetic profile from LN-BA until now!

EurDNA said...

@ Davidski

''You seem to think that you can debunk a linguistics theory with your interpretation of the origins and histories of Scandinavian-specific Y-haplogroups. I'm telling you that you can't.''

It is not my intention to ''debunk'' anything.
Rather I am trying to explain the patterning of data, without claiming any definitive model or truism.

The observations outlined leaves little to personal interpretation, as it merely follows the data (published ancient & modern) and its chronological & contextual progression.

You, on the other hand, are more concerned with black & white positioning.
Therefore, aside from interpretation of the data, the approach is fundamentally different: one is defends one particular model, the other explores the data for what it is.

@ Ric

''It is interesting how some people want to prove that Indo-European families have nothing to do with each other at a Proto Level for some kind of personal and or collective reason...but at the end of the day the fact that PIE existed and dialects and Genes crossing linguistic borders exist doesn't correlate with this Exclusivity Theory on Proto Level. Yes dialect levelling events most probably connected to shared Cultural practices occurred but it does not mean that there were a total break of all linguistic similarities on a dialect level and from the Overarching PIE similarities. If Proto-Indo-Europeans spread as a group into Central Europe from the East, how can there be no connection of their daughter Languages to each other on a Proto Level?''

If that is your interpretation of my ideas, then it is an incorrect one. PIE languages are certainly genetically linked, but their speakers needen't be, but in any case, they still were but not ncessarily in the formulaic way some might envision
Therefor what it entails is analysing each case on its own basis, and this might explain why some adjacent groups might be genetically similar to each other whilst speaking different linguistic families.

EurDNA said...

@ FrankN

''But the process of Germanic differentiating from (late) PIE started long before Grimm's Laws took effect. Pre-proto-Germanic, alternatively also called Germanic Parent Language (GPL), a/o fundamentally reworked the IE tense system''

I agree of course, it was a gradual process.

''Might have been the Nordic BA, might also have been the Nordic MN or LN. ''
Exactly, and in fact, all these changes, without idealising them into discreet entities, were dynamically enough to facilitate more rapid phases of change and shift than background steady state. But, if we are adopting a selectively immobilist viewpoint, then we'd never get to that level of analysis.

''As such, the possibillity of Danish Single Grave adopting TRB language, and only (re-)switching to IE under Unetice influence can't be excluded, albeit it is certainly not the most parsimonous scenario.
On thing, however, is clear: Whatever language merged with late PIE to evolve, as a hybrid, into (pre-)proto-Germanic - it can't have been (Ibero-)Vasconic. The non-IE substrate in Germanic is full of roots that entail consonant clusters, while Basque phonotactics prohibit intra-syllabic consonant clusters.''

That's interesting to know, but I have not envisioned either of those scenarios.

Davidski said...

@EurDNA

It is not my intention to ''debunk'' anything. Rather I am trying to explain the patterning of data, without claiming any definitive model or truism.

Awesome, so can you tell me what you see when you look at the Nordic_LN: RISE98 guy, who is basically identical to many Corded Ware individuals, dated to 2275-2032 calBCE (which makes him the earliest carrier of R1b-U106 on record), and buried in a simple, poor grave?

A member of the Carpathian Basin Bronze Age elite per chance?

Desdichado said...

Granted, Germanic became fully formed in Iron age, but it had been developing in Scandinavia for two thousand years.

Proto-Germanic's ancestor in 1000bc was a very similar language to later Germanic in 100bc. Maybe there was a kind of para-Germanic sphere in Scandinavia in the Bronze age, many related languages related to later Germanic. Later one language replaced the rest.


Chang's linguistic phylogeny from 2015, for what it's worth, posits that Germanic started developing its own separate idiosyncrasies apart from Italic and Celtic somewhere between 2,500 and 2,000 BC, although closer to the more recent end of that range. However, it also has the Germano-Italo-Celtic branch breaking off from the Balto-Slavic a good half millenium earlier, with Indo-Iranian separating from the Balto-Slavo-Germano-Italo-Celtic right around 3,000 BC.

In spite of the phylogeny, though, there was obviously plenty of contact relationship causing borrowings and convergences between Celtic, Germanic and Balto-Slavic. Whether or not a Bell Beaker language superimposed itself over a Corded Ware language in the early Bronze Age and that explains the similarities between Germanic and Balto-Slavic is an attractive model, but must remain speculative. It is obvious, though that Germanic spread from a more constrained geographic area in the centuries immediately prior to its historical attestation, presumably at the expense of Celtic and possibly Balto-Slavic languages that were there already.

EurDNA said...

@ Davidski

''Awesome, so can you tell me what you see when you look at the Nordic_LN: RISE98 guy, who is basically identical to many Corded Ware individuals, dated to 2275-2032 calBCE (which makes him the earliest carrier of R1b-U106 on record), and buried in a simple, poor grave?

A member of the Carpathian Basin Bronze Age elite per chance?''


Nope. U106 can hypothetically be broadly linked with the Dagger Grave horizon which expanded from Denmark toward the north c. 2200 BC. Not much data to go on though, but an interesting possibility.

Davidski said...

@EurDNA

Nope. U106 is broadly linked with the Dagger Grave horizon which expanded from Denmark toward the north c. 2200 BC.

What do you mean by "broadly linked"? You and your friends linked him to it?

jan.t.andersson said...

@Davidski, Bob Floy, weure:

Ruling the world was meant to be a joke - hitchhiking with the English and then American feats around the globe. Or crimes, if you so please. I'm perfectly happy ruling just a few blocks in Stockholm.

Besides that, it seems odd to make "Germanic" the center of gravity when discussing these cultures. So NBA is pre-germanic. And FDC is pre-pre-germanic. And hällkistetid is pre-pre-parallell-germanic? Though battle axe is definitely pre-pre-pre-germanic. What about TBK? It's not germanic at all but we should still define it in relation to everything's focal point. Pre-pre-pre-pre-not-germanic, then.

Tacitus' "Germania" has been called the most dangerous book ever. So let's now ditch that term altogether.

weure said...

@Jan It ain't German until the fat lady sings a Schlager ;))))

Desdichado said...

Tacitus' "Germania" has been called the most dangerous book ever. So let's now ditch that term altogether.

No. It's convenient, accurate, and everybody knows what it means. Just because there are disputes about exactly where it begins doesn't take away from it's incredible utility.

Samuel Andrews said...

@jan.t.anderson,
"Besides that, it seems odd to make "Germanic" the center of gravity when discussing these cultures. So NBA is pre-germanic. And FDC is pre-pre-germanic. And hällkistetid is pre-pre-parallell-germanic?"

Anyways, the reason commentators care are mentioning Germanic is because the origin of Indo European languages is the focus here. Proto-Germanic is an Iron age language but it had been developing in northern Europe for two thousand years before that. Tracing back its ancestors in those two thousand years is relevant to origin to the origin of this huge modern IE language branch.

Tracing back the distant origins of all major modern IE branches no matter helps trace back the origins of Proto-Indo European the mother of all of them. It's the same reason people here link Bronze age/Corded Ware Baltic to modern Baltic & Slavic languages. Or link, Romance/Latin to Bronze age Italy.

weure said...

@Sam of course but that doesn't mean that in the whole area where we find that LN-BA genetic pattern (that last until todday), has also the same kind of unity in the language development..

Samuel Andrews said...

@jan.t.anderson
"Tacitus' "Germania" has been called the most dangerous book ever. So let's now ditch that term altogether."

Ditch the term Germania?

Tacitus didn't create the term Germania. Ceasar used this term as well. He noted, that the Germans had distinct language & customs from the Gauls he was at war with.

The term, Germania, was an accurate term used in Roman period Europe to describe an ethno-lingustic-cultural region. Germania basically became modern Germany the country. Of course how this happened is complicated.

Ceasar's conquest of Gaul was the first step in creating the modern boundary between France & Germany. The last step was 9ad in the battle of Tuetonburg forest. This gurenteed Rome would not invest in conquering Germania & there enshrined the the border between Romance speaking France & Germany.

None of this is dangerous information. People can use it in dangerous ways but the bare facts about it are not dangerous.

There has not yet been a thorough genetic analysis of Germany. It is quite possible, eastern Germany (which was Slavic land in Middle ages?) is heavily Slavic in ancestry. It is quite possible, western Germany is heavily French in ancestry. Whatever the case, the reality of Germania as becoming Germany is real.

Samuel Andrews said...

@weure,
"@Sam of course but that doesn't mean that in the whole area where we find that LN-BA genetic pattern (that last until todday), has also the same kind of unity in the language development.."

Totally Agreed.

Andrzejewski said...

@Saumuel Andrews ". It is quite possible, western Germany is heavily French in ancestry. Whatever the case, the reality of Germania as becoming Germany is real."

LOL...actually I strongly believe that it's vice versa, the other way around: with the Franks' conquest of Gaul and their heavy settlement in Northern France north of the Loire River, the Viking raids on Normandy and the German-speaking Alsace-Lorraine inhabitants west of the Rhine, it's much more likely that the French are very German-shifted. TBH, it amazes me how come the French don't speak a Teutonic language by now.

epoch said...

@jan.t.andersson

"Ruling the world was meant to be a joke "

Let's hope the rest was meant as a joke as well.

Bob Floy said...

@jan.t.andersson

"Tacitus' "Germania" has been called the most dangerous book ever. So let's now ditch that term altogether."

Jesus Christ you people need to calm down.

weure said...

Some add as North Dutch we are in the Nordic LN-EBA or 'proto-Germanic' corner mom and I are 'twins' of Rise71 and Rise97.

What causes the similarity? Rise71 and Rise97 are both LN-EBA samples of Eastern Denmark. This area is known as TRB stronghold, even after the incoming Single Grave people. That's also in Drenthe the case: a TRB hotspot with incoming Single Grave people.

In LN-EBA we see a 'de-neolithisation', see: Rune Iversen, Beyond the Neolithic transition - the ‘de-Neolithisation’ of South Scandinavia (2018).

Keynote in this respect:
"In the eastern parts of Denmark (incl. easternmost Jutland) and Scania, the late Funnel Beaker culture continued until c 2650 BC with some regional developments in pottery style. Megalithic tombs were the predominant burial form throughout the Neolithic. Already from c 3000 BC Pitted Ware elements were incorporated in the gradually disintegrated Funnel Beaker tradition. The reuse of megalithic tombs is in particular visible on the Danish Islands where Pitted Ware tanged arrowheads and certain variants of Single Grave battle-axes and beakers were incorporated in the existing burial custom illustrating the gradual disintegration of the Funnel Beaker culture. In Scania and on Bornholm, the Swedish-Norwegian Battle Axe culture constituted a particular variant of the widespread Corded Ware complex."

Still find it astonishing that even after 4000 years the genetic profile shows a similarity with that of LN-BA.....

jan.t.andersson said...

@Samuel Andrews:

No, ditch ”Germanic”, especially when its used for like anything Scandinavian or with roots in Scandinavia. Gaius Julius considered some Gallic tribes to be Germanic. Needless to say, he was a politician and a general, not an anthropologist, and defined peoples to fit his goals. Why even consider him. ”Germanic” culture and language developed and matured in Scandinavia so why not name them Scandinavian.

BTW: do you perhaps have post-Scandinavian ancestry?

Matt said...

Proto-Germanic regions of the IE dialect continuum being confined to southern Sweden and Denmark and then expanding southwards into Northern Germany from there is really just a supposition, and not really a very good or solid grounds to go and rename a branch of Indo-European with a reasonably good name as it is.

jan.t.andersson said...

Check out the graphics in the following text about a TV program broadcasted a while ago. Comparing a couple of battle axes with the program host. Continuity, it is.

https://www.svt.se/nyheter/vetenskap/ny-dna-forskning-visar-att-svenskar-framforallt-harstammar-fran-ukraina

old europe said...

jan.t.andersson

can you provide a summary of the article

Bob Floy said...

@jan t. andersson

It may come as a shock to you that some of us don't speak Swedish.

Matt said...

Few graphics to try and visualize continuity and change in Northern Europe: https://imgur.com/a/JyL4EUP

Continuity probably fairly high, at least considering in terms G25 measures. Sweden is quite diverse so continuity probably less there than Denmark / Norway? May be some tug of war between genetic influences into region from West and East. But this is probably geographically strutured - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5765326/.

Gabriel said...

@jan.t.andersson

So I guess Germany is now Scandinavia?

Just because Tacitus was a politician and not a anthropologist and because he called potentially Gaulish tribes Germanic doesn’t mean that we have to stop saying “Germanic”. We can’t just be so corrective.

Suyindik said...

@Davidski

As Ryukendo from anthrogenica pointed out, there were only 12-13 samples from the Iron Age period of Latium in Italy. Lets say that half of these samples are "males", meaning there are only 5-6 Iron Age Y-DNA haplogroups available. Do you know the Y-DNA haplogroups of these? Or do you only have information regarding the autosomal ancestry? How many Y-DNA are there of Etruscans, and do they provide enough historical and archaeological evidence for the defining of which individual is Etruscan or Italic?

Ryukendo from anthrogenica pointed to "EHG", "Levant_N", "Iran_N" ancestries in the Iron Age period. You are saying that the Etruscans are a two-way mix of "early farmers" and "Bell Beakers", plus maybe some "Roman input". Do you base this conclusion on autosomal ancestry data or Y-DNA data?

Historical sources called the Italic people aborigines of Latium. If we associate "EHG" to the Italic people in Latium, and associate "Levant_N" and "Iran_N" to the Etruscans from Tuscany, could it mean that the immigrant Etruscans, settled first in Tuscany, and then expanded into Latium and in Latium the Etruscans mixed with the aborigines(Italic people)? Because there are lots of historical sources showing marriages between Latin and Etruscan royals.
Could it be that if in the future Etruscan burials from Tuscany will be studied, these will be having a lot of "Levant_N" and "Iran_N" Y-DNA?

Davidski said...

@EurDNA

You're the one who started with the LOLs and sarcasm. If you can't hold a debate without getting snooty, then don't post here.

Davidski said...

@Suyindik

I don't have any info about the Y-haplogroups of the early attested Italic speakers and Etruscans, but they're a fairly homogeneous bunch, and very likely to be overwhelmingly a two-way mixture of Central Europeans rich in steppe ancestry and local EEF groups.

I don't know if the Italic speakers and/or Etruscans have any extra Iran_N/CHG that isn't part of their steppe ancestry. Based on what I've seen, I'd say no. But even if they do, then it's probably very minor.

It should be interesting to see how the authors of the paper(s) explain the obvious genetic similarity between the Italic speakers and Etruscans, but to me it looks like the Etruscan language was adopted by some Italic speakers without any significant accompanying gene flow from outside of Iron Age Italy.

The Roman period is when things get a lot more interesting, with the genetic homogeneity of the Iron Age disappearing as a result of gene flows into Italy probably from practically all directions, but mostly the eastern Mediterranean. This process results in a genetic Roman cline running from the former Iron Age cluster to the eastern Mediterranean. Part of this cluster overlaps with Mycenaeans, but most of it closely resembles the modern Italian cluster, except it eventually extends further south into the Near East.

Bob Floy said...

I wonder how the Villanovan culture fits into it?
The Etruscan civilization appears in almost exactly the same spot, just as that culture ends. And some Villanovan pottery motifs look an awful lot like Sintashta/Andronovo, so it's hard to imagine that they weren't indo-European. Hmm.

FrankN said...

Bob: I'd say Villanova is kind of Iberian BB reloaded: They were originally IE - obviously, since Villanova emerged out of Urnfield. Some switched to pre-existing non-IE - Etruscan, Rhaetic - in the process possibly overforming the former. Others stayed IE but acquired local sub-/ adstrate: The sound shifts defining proto-Italic, PIE bh->f/v, PIE gh->h, obviously reflect Etruscan influence. The Etruscan Alphabet neither has a letter "b", nor "g", so both sounds were obviously alien to Etruscans and accordingly had to be shifted in order to allow them speaking IE (Italic).

The more interesting question is whether IE was already present in Italy before the Urnfield/Villanova expansion. The MLBA Appenine culture displays heavy Aegaean influence, and it is tempting to connect well-known ItaloCeltic-Anatolian isoglosses such as the formation of the passive to MLBA Anatolian settlement of S. Italy. I mean, Livy and Virgil state that Romans originate from Troy. Of course, J. Ceasars suggested paternal descent from Aeneas was a myth (or political propaganda), but sometimes, founding myths hold a grain of truth..

Andrzejewski said...

On the linkage between Etruscan and Kartvelian:

https://www.academia.edu/35555751/A_structural_comparison_of_Etruscan_with_the_Kartvelian_languages

EurDNA said...

@ FrankN


From what I recall, there is a big gap in the sample set (from the circulating rumours) between early Bronze Age & Iron Age to Republican Period. The Bronze Age genomes are said to be predominanly EEF/ WHG - in other words we can expect them to a continuation of Remedello-like groups.

Following on from this, we should not forget the Cetina culture expanding from the East Adriatic, which made a significant impact in eastern Italy & Peloponessus; which might account for some of the Aegean influences you mention (in fact, possibly a confounder, because the influence/ migration went Cetina -> Italy; Cetina -> western Greece); whilst true Aegean / Mediterranean impact I'd imagine limited to southern Italy & Sicily.
In this regard, the Polada culture, poor in anthropological remains, will hopefully get some coverage in the future. It will be important to clarify how much continuity there is between it & preceding BB horizon in Italy, and what to make of the emergent elites with links to the Balkans there. Indeed, to impress again, c. 22/ 2000 BC seems to be a formative moment for central Europe, when previously Late Neolithic tribal societies evolve into something more connected & sophisticated whislt largely maintaining their genetic backgrounds.

Matt said...

Re; Etruscan and its place in the world's language, typological and semantic clustering schemes can give strange results and place languages which do not have family relationships close together through chance or influence:

1. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0205313 - "Diachronic Atlas of Comparative Linguistics (DiACL)- A database for ancient language typology". See clustering in Fig 6

2. https://arxiv.org/pdf/1904.10820.pdf - "Semantic Drift in Multilingual Representations". See Fig 4.

You can get all kinds of non-family relationships - e.g. Basque with Caucasus languages (including Kartvelian languages) in 1 and English with Romance rather than Germanic, and Romani (Indo-Aryan) with Slavic in 2 (rather than Romani with other IA languages). That is apparent even with the sets above, where they're dealing with pretty well attested languages biased to be heavily from one family.

Phonology (and features like linguistic tone, etc.) gives similar results with non-family connections.

The only way to reconstruct relationships is through reconstruction of basic lexicon as cognate through reconstruction of sound change, supplemented with shared grammar (not typology). That's agreed upon within linguistics, with the debate really then about the best way to then proceed once families are established through this method (is it better to try and put sound and grammar changes into some sort of tree, or is this approach just so plagued by tree inconsistent sharing or lack of tree like structure that sheer quantitative counting of lexical sharing works better?). So far, under this approach, Etruscan, Iberian, Vasconic and so on are all still isolates.

Davidski said...

@EurDNA

From what I recall, there is a big gap in the sample set (from the circulating rumours) between early Bronze Age & Iron Age to Republican Period. The Bronze Age genomes are said to be predominanly EEF/ WHG - in other words we can expect them to a continuation of Remedello-like groups.

Italy_EBA forms a cline from Northern Europe to EEF. Some of these samples have a lot of steppe ancestry.

Early Italic speakers, from Central and South Central Italy, and Etruscans cluster together along this cline, right near Iberians but a little closer to EEF. So they also have a lot of steppe ancestry.

None of the Romans cluster in this Iron Age space; they're all shifted south much closer to the Near East.

These aren't rumors.

Bob Floy said...

So the Romans look like southern Italians, in other words.

EurDNA said...

@ Davidski

''Italy_EBA forms a cline from Northern Europe to EEF. Some of these samples have a lot of steppe ancestry.

Early Italic speakers, from Central and South Central Italy, and Etruscans cluster together along this cline, right near Iberians but a little closer to EEF. So they also have a lot of steppe ancestry.''

Okay. I thought I had read that Steppe ancestry arrives later, like MLBA. It seemed odd at the time, perhaps I misunderstood it.

Davidski said...

@Bob Floy

So the Romans look like southern Italians, in other words.

Yes, but their cluster is more heterogeneous, and reaches further north and south, as far south as Cyprus. A minority of the Romans are also obvious outliers and cluster with various Near Eastern groups.

Arza said...

@ Davidski
"Some Romans will be basically indistinguishable from Greeks, but this is unlikely to be the majority of them."

From modern Greeks?

Bob Floy said...

Odd, considering that Latin is more conservative than "P" Italic languages like Oscan and Umbrian. I would have expected things to be the other way around, with groups like the Samnites plotting closer to where the Romans are, and vice versa.

Davidski said...

@Arza

I actually meant Mycenaeans in that comment, rather than modern Greeks.

@Bob Floy

Samnites plot far to the north of Romans, right near and even among Basques.

These Iron Age Italic groups are homogeneous, which isn't surprising. Romans are very mixed, which again isn't surprising, because much of Rome was located outside the Italian Peninsula.

Bob Floy said...

@David

"Romans are very mixed, which again isn't surprising, because much of Rome was located outside the Italian Peninsula."

Ok, for some reason I was thinking they were earlier Romans.
It would be nice to see what some true Latins looked like, I guess they'd probably look more like the Samnites and Umbri(and Etruscans).

Davidski said...

I'm pretty sure that the earliest Romans would indeed cluster with the Umbri, Etruscans etc.

I doubt that any of the currently sampled Romans are from the earliest times of Rome.

EurDNA said...

Ah , so i remembered correctly
Having tracked down the original message from our kind messenger:

UPPER PALEOLITHIC
All WHG

NEOLITHIC
Mostly EEF, some WHG. Some Iran_N, quite a significant quantity, as much as WHG. PCA position Between Sardinia and Maltese, east of Sardinia,closer to Sardinia than to Maltese. Very homogeneous.

BRONZE AGE (EARLY)
Overlaps modern-day Sardinia, Iran_N percentage declines, WHG and EEF increases
Very homogeneous.


IRON AGE TO REPUBLICAN PERIOD (700-20BC)
Note: Separated from previous period by 1000 year gap.
Fewer samples, of those that exist 60% overlap with North Italy, 40% overlap with South Italy and Sicily, centroid of overall cluster in central Italy but no samples occur there, very wide spread.
EHG appears
Levant N Appears for the first time, sporadic and inhomogeneous distribution, Iran_N increases further.

IMPERIAL PERIOD
Dense cluster centroid between Greeks, Cypriots, South Italians/Sicilians, and Syrians, closest to Sicilians. Long tail stretching from central cluster to Syrians and Iraqi Jews. Couple of Northern-shifted samples overlapping N Italy, France, Spain.
Iran_N increases further, Levant N again sporadic and inhomogeneous.

LATE ANTIQUITY
Tight cluster centroid in S Italy, in the same place as in the previous period. Southern tail to Middle East disappears. N Italian, Northern European and NW European outliers exist.

AFTER
Resemble modern central Italians.

If this was the same study, then we should have early Romans, and Italy was already ''diverse'' pre Empire

Davidski said...

That's from a different paper than the one I've got the PCA from. I guess we can put the differences down to a different sampling strategy.

EurDNA said...

That's awesome. So 2 great papers; and with more in the future, sampling strategy will no longer be an issue (as it arguably has thus far).

Suyindik said...

@Davidski
When will both studies be published? Will it become like the South Central Asian study which was supposed to be published a couple of months ago?

Davidski said...

I don't know. Maybe soon, whatever that means.

jan.t.andersson said...

@Bob Floy:
"It may come as a shock to you that some of us don't speak Swedish."

Since I know for a fact that most doesn't understand Swedish, I'm not shocked at all. Though it puzzles me. But it's kind of a shock to learn that some people doesn't know about Google Translate.

BTW: Are your ancestry perhaps also post-Scandinavian?

@old europe:
"can you provide a summary of the article"

Since you ask politely:

The program host had his DNA compared to two 4500 year old battle axe individuals from a grave called Bergsgraven in Linköping. That comparison is what is shown in the graphics: A small part hunter/gatherer, a bit more of early farmers, most of it Yamnaya. Same pattern 4500 years ago as today. For geneticist Helena Malmström, Uppsala university, this is not surprising: "All people with roots up here are quite similar."

jan.t.andersson said...

@Gabriel:
"So I guess Germany is now Scandinavia?"

No. Geographical areas don't move. Culture and people move. Like when the culture, language and people spread southwards from Scandinavia into the continent - south, west and east. Not only present-day Germany. Like all post-Scandinavian peoples, Germans may call themselves and their country whatever they like.

Someone tried to justify "Germanic" by reference to some old Romans. It may be true that just because they got it wrong doesn't mean we should stop using it, but it doesn't mean we should stick to it either. The old Romans are simply not relevant. What is relevant when naming something is its character. The characteristics of this culture and language and people is that it developed and matured in Scandinavia for something like 2000 years before spreading. Then, should we perhaps call it Scandinavian?

old europe said...


@jan.t.andresson

I think mainstream linguistic clearly sees the scandinavia/northern Germany region as the most likely "germanic homeland". So your point is not that wrong even tough there are nowadays a minority of scholars that see the Harz region as a likely homeland too. But I do not think that the british and the "celto-germanic" like the bavarian the swiss and austrian would easily agree to this change of nomenclature.

@dave
Thanks for the translation. It intrigues me that they refer to the source of the migration as Ukraine ( even tough they use the outdated Yamnaya nomenclature). Maybe they are hinting at Sredni Stog as the origin of CWC.

weure said...

@Jan the point is that I have also as you call it" Scandic genetic" profile, at least it's congruent with Nordic LNBA

Still it's not due to a Scandic migration, but due to a combination of TRB and Single Grave. So what solution it will be if we replace Germanic by Scandic? Ok Germanic is a Roman invention. But Scania too:
Wiki: "The name "Scandia", later used as a synonym for Scandinavia, also appears in Pliny's Naturalis Historia (Natural History), but is used for a group of Northern European islands which he locates north of Britannia. "Scandia" thus does not appear to be denoting the island Scadinavia in Pliny's text. The idea that "Scadinavia" may have been one of the "Scandiae" islands was instead introduced by Ptolemy (c. 90 – c. 168 AD), a mathematician, geographer and astrologer of Roman Egypt. He used the name "Skandia" for the biggest, most easterly of the three "Scandiai" islands, which according to him were all located east of Jutland."

So yes the genetic profile is Scandic but it's broader than Scandic. Is it Nordic yes it's Nordic but then comes the word Nordicist around the corner. Pffff

So with all the pro's and con's Germanic is IMO still the most useful. Because in the common opinion are all Scandinavians Germanic but not all Germanics are Scandinavian. And in this case the most Scandinavians have a LNBA Nordic genetic profile but not all with a Nordic LNBA genetic profile are Scandinavians....But I guess it's defendable to say that most people with a Nordic LNBA genetic profile are Germanic as it is accountable for Scandinavians, North Germans and North Dutch.....

Bob Floy said...

@jan. t. andersson

"it's kind of a shock to learn that some people doesn't know about Google Translate"

Can't be bothered in this case.

"Are your ancestry perhaps also post-Scandinavian?"

No, nor does one need to have ancestry from a particular place to take an interest in it.

Bob Floy said...

@EurDNA

Thanks for that summary.

So Italy got a "quite significant" dose of Iran_N in the neolithic.
Interesting.

Matt said...

@Davidski: Early Italic speakers, from Central and South Central Italy, and Etruscans cluster together along this cline, right near Iberians but a little closer to EEF. So they also have a lot of steppe ancestry.

On a tangent to this, I feel like I am again obliged to complain again* about how those genetic isolates in North East Italy haven't been looked at again in light of ancient dna - https://www.nature.com/articles/ejhg2012229 - "Genetic characterization of northeastern Italian population isolates in the context of broader European genetic diversity"

PCA positions in a European PCA: https://media.nature.com/lw926/nature-assets/ejhg/journal/v21/n6/images/ejhg2012229f2.jpg

PCA positions in a West Eurasian PCA: https://media.nature.com/original/nature-assets/ejhg/journal/v21/n6/extref/ejhg2012229x3.pdf

Many of them totally compatible with being clinally intermediate Sardinians and Basques in PCA with European and West Eurasian subjects. Suggests that these small villages may well be effectively preserved Iron Age Italian/North Italian populations.

*Basically in the vague hope that some researcher will think my grumbling is on to something and try and get a paper written, if there is anything!

FrankN said...

Dave: How do Nordic_LN and Nordic_BA samples compare to BB_NL?

I mean: If

(a) Nordic_LNBA genetically reflects SingleGrave - TRB/GAC interaction (with predominance of the former, but steady increase of the latter over time),

(b) this interaction resulted in the emergence pre-Germanic as IE family with considerable divergence from PIE with respect to basic vocabulary and morphology (esp. the verbal system),

(c1) Dutch SingleGrave wasn't that different from Jutland, encountered also TRB substrate, and after ca. 2350 BC was connected by strong economic ties (flint dagger import) to the Danish LN Flint Dagger Culture, and
(c2) Nordic_LN clusters with modern Dutch,

shouldn't we expect late Dutch Single Grave, and by extension Dutch_BB, to have spoken something akin to Jutlandish pre-Germanic?

Or was there some extra genetic component incorporated in Dutch_BB to sufficiently set them apart from Nordic LNBA to suggest their language could have been quite different? If so, which extra influence would we talk about.

The question may be extended to N. England (I6679, I2618, I2609, I1767, I5382, I7635) that also substantially imported Danish Flint Daggers and during the EBA used Halberds very similar to Danish ones (in this case, the flow is assumed to be from England into Jutland).

Finally, Irish/ Scotch/ N. English Food Vessels bear intriguing parallels to Schönfelder pottery from the Magdeburg area. The Quedlinburg_BB->Rathlin connection has already been discussed in length. How does N. English I2421, Food Vessel, 1931-1756 calBCE compare to both?
https://en.wikipedia.org/wiki/Food_Vessel
https://www.scottishheritagehub.com/node/1521 [see esp. Fig. 73 and compare it to the Schönfelder bowl here:]
https://st.museum-digital.de/data/san/images/201407/25192949378.jpg

weure said...

@Frank,
Ad c not to forget that TRB West may have been a derivative of TRB North. See your own Tiefstich expansion. the link with the red Helgoland flint very often found in Drenthe. And TRB West has more Ertebølle features than Swifterbant features,
So when North Dutch Single Grave and TRB Show affinities with the Denmark versions.....no wonder that the genetic similarity is big.

Andrzejewski said...

@All How safe is it to assume that the reconstructed PIE language was actually the one spoken by the CWC folks rather than the Yamnaya proper, based on the fact that most IE daughter languages (except for Anatolian branch and perhaps Tocharian) derive from CW? CW/Battle Axe -> Single Grave Culture -> Bell Beaker?

Davidski said...

Tocharian samples from Silk Road sites are on the way.

What if they're packing a lot of Sintashta ancestry and R1a-Z93?

Andrzejewski said...

@Davidski "ocharian samples from Silk Road sites are on the way.

What if they're packing a lot of Sintashta ancestry and R1a-Z93?"

It wouldn't surprise me: If Tocharians = Tarim Basin Mummies = Yuezhi/Kushan, it turns out that the TBM are closer to Andronovo horizon than to Afanasievo. However, due to the settlement of Saka in the Tarim Basin and the silk road transfer of Buddhism associated with them, there must've been a lot of linguistic assimilation between those two Indo-European groups. One thing is for sure: both according to depictions done by the Chinese and also judging by the appearance of the Tarim Basin Mummies -> those Tocharians/Saka/Wusun groups looked very European, even by modern standards.

Andrzejewski said...

Going off on a tangent re: Tocharians: there was evidence that many features of the nascent Chinese Han civilization were developed as either an influence of or as a counter-reaction to the invasion of European groups: Chinese mythology resembles Indo-European ones; words like "dog", "honey" and many others ("horse" and "chariot", of course) are originally Indo-European; metallurgy and even Asian martial arts were developed locally to repel the invading IE's; and of course - Buddhism was later transmitted to China from the Andronovo-derived culture of post-Harappan Indus Valley Northern India (and Nepal) via the silk road by Tocharian/Kushan/Yuezhi merchants.

EurDNA said...

@ Andrzejewski

''How safe is it to assume that the reconstructed PIE language was actually the one spoken by the CWC folks rather than the Yamnaya proper, based on the fact that most IE daughter languages (except for Anatolian branch and perhaps Tocharian) derive from CW?'''

How would Italic, Albanian or Greek relate to Corded Ware ?
Outline your model here. Which are the Italic & Balkan branches of R1a

Davidski said...

@EurDNA

Outline your model here. Which are the Italic & Balkan branches of R1a.

But is this question relevant?

Bell Beakers with typically Corded Ware genetic structure entered the Carpathian Basin from the north or northwest during the Copper Age and here they met and mixed with the rather different late Yamnaya.

Where did these Bell Beakers come from?

Davidski said...

Here's a clip on the topic that I've posted in the comments in the past...

https://www.youtube.com/watch?v=cBZbt4mNb7U&t=12s

Bob Floy said...

@Davidski

"Tocharian samples from Silk Road sites are on the way.

What if they're packing a lot of Sintashta ancestry and R1a-Z93?"

Then Dragos will have been right, heheh.

Davidski said...

Maybe he already knew?

Bob Floy said...

Well he does like to play "I know something I won't tell".
Except he really does know, unlike Nirjhar and his ilk. They don't even know things that were revealed years ago.

EurDNA said...

@ Davidski

''Bell Beakers with typically Corded Ware genetic structure entered the Carpathian Basin from the north or northwest during the Copper Age and here they met and mixed with the rather different late Yamnaya.

Where did these Bell Beakers come from?'''


Hungarian BB seems fairly heterogeneous autosomally, but nothing suggests a particularly heavy CWC adstrate. Their uniparental markers, being H2, I2a1a, I2a2a1b, G2, R1b-Z2103 certainly don't point to CWC; even if we accept that R1b-51 came from SGC (which I don't, because Bb & CWC are separate entities)
In any case, BB isn't particularly relevant for Balkan Bronze Age, given that it had essentially disappeared by 2400 BC from Hungary.
BB's role for Italic remains to be evaluated.

The southernmost extent of CWC is the Slovak-Nitra group.

Ric Hern said...

Their Ancestors came from the Northwest Block...

Samuel Andrews said...

@Matt,
"On a tangent to this, I feel like I am again obliged to complain again* about how those genetic isolates in North East Italy haven't been looked at again in light of ancient dna - .......
Many of them totally compatible with being clinally intermediate Sardinians and Basques in PCA with European and West Eurasian subjects......
...*Basically in the vague hope that some researcher will think my grumbling is on to something and try and get a paper written, if there is anything! "

Yeah. Maybe, eventually their DNA will be published. Northern Italy in general seems to be a continuation of Bronze age/Iron age Italy. From what David says, it sounds like Etruscans & Italic tribes cluster in northern Italy.

Samuel Andrews said...

@Davidski, Are you saying Etruscans & Italic tribes cluster with modern northern Italians. Roughly 40/60 Bell Beaker/Neolithic Italy. Or are they much more Neolithic Italy.

Based on modern DNA, it looks like the Italian cline is a mix between northern Italy & Greek Islanders (ancient western Asia Minor?). So, it would make sense Iron age Italy cluster with modern northern Italy.

Davidski said...

@EurDNA

That video I linked to talks about female exogamy between typically brachycephalic Beakers and Carpathian Basin locals. And the Hungarian Beaker with Z2103 looks like an obvious first generation mix between Central European Beaker and Yamnaya.

Also, one of the Hungarian Beakers does belong to the typical Beaker P312.

So there's now direct evidence of Beakers being intrusive into the Carpathian Basin, probably from way up in the north, and mixing with Yamnaya there.

Of course, Hungarian Beakers do show significant ancestry typical of the Corded Ware complex. They can even be described as a three-way mix between Corded Ware, Hungarian Neolithic farmers and Yamnaya.

But yeah, we can try to be more precise and say that they look like a mix between (the very brachycephalic) Rhenish Beakers, Hungarian Neolithic farmers and late Yamanya. This model is basically the same thing though.

I don't know how the Slovak-Nitra group fits into this, but what I do know is that the Hungarian Beakers show a very strong signal of ancestry from near the North Sea.

Samuel Andrews said...

@jan.t.anderson,
"Someone tried to justify "Germanic" by reference to some old Romans. It may be true that just because they got it wrong doesn't mean we should stop using it, but it doesn't mean we should stick to it either. The old Romans are simply not relevant."

"German" just happens to be the name that became popular to describe all the similar tribes who lived east of the Rhine river. Tacitus, who never visited Germania, gives a story for why this name was chosen. There was tribe with that name, then it became the namesake for all of them.

It is weird that now, linguists put Norse languages under the same "Germanic" label even though Scandinavians were never called German by anyone. It is the same with how Gealic is called a "Celtic" language because their lingustic relatives in mainland Europe were all called "Celts" but Irish were never called Celts.

Bob Floy said...

@Sam

IIRC there was one particular tribe which the Greeks referred to as the "Keltoi", and that became a shorthand for the entire ethno-linguistic group. So it's exactly the same thing as with the Germanic folk.

jan.t. andersson

With that in mind, who cares? What the hell is the big deal?

Davidski said...

@Samuel Andrews

Are you saying Etruscans & Italic tribes cluster with modern northern Italians. Roughly 40/60 Bell Beaker/Neolithic Italy. Or are they much more Neolithic Italy.

Etruscans and Italic tribes cluster next to Iberians and North Italians, but they're a little more shifted towards EEF.

So they might have almost as much steppe ancestry as most Iberians, or as much as Iberians, if the PCA plot doesn't tell the whole story, and it might not, because it just shows the first two dimensions.

weure said...

@Bob totally agree....the big deal is that the names for this 'genetic cline' are in the past misused by people with a deviation in the right arm. The blonde Aryans...and such like debunked things....

Bob Floy said...

@Weure

I know, believe me, haha.
My real point is that it's time to move on and not throw the baby out with the bathwater.

Matt said...

Samuel: Yeah. Maybe, eventually their DNA will be published. Northern Italy in general seems to be a continuation of Bronze age/Iron age Italy. From what David says, it sounds like Etruscans & Italic tribes cluster in northern Italy.

Maybe they will, it would be good! I'm not sure it ever will be as my guess would be that there were some legal matters which prevent wide release in that project. It's not exactly hugely important (and we will know the true history of Northern Italy anyway through ancient dna) but it would be cool to discover if that were true or not.

Andrzejewski: here was evidence that many features of the nascent Chinese Han civilization were developed as either an influence of or as a counter-reaction to the invasion of European groups

There is much discussion of this kind of area by Victor Main and "Sino-Platonic Papers" if you wish to read more. My view is that judging by the weak evidence of Western steppe ancestry at DSK Eastern steppe, any influence in early stages of Chinese civilization was probably indirect and there were probably never many IE speakers in direct contact with North China, much less any invasion. Chariots and horses spread to China probably only by intermediaries who were not IE speaking. I would also doubt that the Chinese word for dog is a loan from Indo-European, though the others may be. However:

even Asian martial arts were developed locally to repel the invading IE's

is pure WTF. Chinese have a tradition of martial arts because virtually all cultures have formalized ways of teaching wrestling and boxing. Nothing to do with IEs. They also have a unique literary tradition unique to them inspired by legends of Buddhist and Taoist monks doing great feats, due to their enlightenment, where they learn secret special martial arts styles inspired by animals and so on. Again nothing to do with invading IEs and only to do with IE speakers at all via Buddhism and India.

Suyindik said...

@Davidski

And regarding the Y-DNA, what percentage do you give for the EEF haplogroups among the Etruscans? And based on how many samples is this? From which regions? Are these different regions from the Moots study from Stanford?

And I must say that in the past 2 years I have seen dozens of pre announcements of unpublished studies in forums and blogs, and only vague descriptions are revealed, no real data is leaked. And in the end the study wont be published for years(while promising to publish soon in a couple of weeks), and will probably not be published in 100 years.
Even for the Damgaard et al ‎2018 study, it was speculated for years(and it was always going to be published soon) and then they actually publish only 25% of their Y-data, the rest is assembled by volunteers on the internet.
What do you think about this matter, is this related to political issues? How can science be done when politics is involved? Why are scientific groups not obligated to give a single precise date of publication? Why is it not forbidden for scientists to reveal leaks(without actually giving the real data) for the gain of attracting people and sponsors towards themselves?

Davidski said...

@Suyindik

I don't have any info about the Y-haplogroups of the ancient Italian samples.

But I'm pretty sure that this paper will be published soon, probably this year that is, and all of the samples will be released as well.

Gaska said...


@Davidski, EurDna

A good number of Czech and Hungarian mitochondrial haplogroups are typically Iberian, and some Hungarian BBs have good percentages of Iberian chalcolithic. Trying to make the CWC part of the BB migrations does not make much sense (where is R1a?). Simply the Eastern BBs stopped in Hungary the Indo-European migrations from Yamnaya, even mixing with them. Regarding the participation of the CWC in all this history, is limited to provide a strong steppe signal in the BBs outliers of the North (Saxony, Frisia, Denmark)-In any case I am happy to see that many people agree that the Eastern BBs are not horsemen of the steppes, but Western/Central BBs. 5-6 months ago when I said it in Anthrogenica, only a few believed it, others intelligently, took note and moved their positions

Regarding Italy, I do not know if the geneticists are going to be able to establish a genetic continuity as clear as that existing in Iberia between the Chalcolithic and the historic peoples of the Iron Age - We have Bbs in Sicily (2,300 BC) and Parma (2,065) BC) that are also very Iberian (maybe that's why the Etruscans and northern Italy resembles the Basques and northern Spain) but nobody at the moment is able to relate them with later Italian prehistoric cultures.As indicated by the uniparental markers, the CWC did not participate in the expansion of the IE languages ​​in southern Europe, I think they are much more modern (Iron Age) than the BB culture.

EurDNA said...

@ Davidski

''but what I do know is that the Hungarian Beakers show a very strong signal of ancestry from near the North Sea.''

What exactly do you mean by that ?
Can you provide a model ?

Bob Floy said...

@David

According to Raveane et al., "Iran Neolithic (IN) ancestry was detected in Europe only in Southern Italy." Why would that be? Do you think it's just a result of back-and-forth between the Roman empire and it's near eastern possessions?

Davidski said...

@EurDNA

Here's a model of the P312 Hungarian Beaker as Dutch Beakers + Baden farmers. Can't remember seeing a better fit in a model with so many outgroups and 400+ SNPs retained.

https://drive.google.com/open?id=1mP7tde-qh281QCqJYg6q5y5Wx9OZe8V3

Here's a model of the Z2103 Hungarian Beaker as Dutch Beakers + Yamnaya. Works well enough and basically comes out 50/50.

https://drive.google.com/open?id=1qcq4spxzyyUTF_DD95H3sIUHjYx0JcBi

And here's a PCA that shows why these models work so well.

https://3.bp.blogspot.com/-D66Bv_G8j1c/XFyKNL48eOI/AAAAAAAAHjQ/YGYX4_0pLJE82G4pRoziypbS4wrHJpIAACLcBGAs/s1600/Mapping_Beakers_PCA.png

Davidski said...

@Bob Floy

Do you think it's just a result of back-and-forth between the Roman empire and it's near eastern possessions?

Probably, because in the PCA that I've seen Romans aren't only shifted far to the south of the Iron Age Italic groups, but most of them are also sitting clearly east of the Mycenaeans.

Bob Floy said...

@Dave

So southern Italy then had some serious input from places like Syria, Anatolia, etc., in late antiquity or thereabouts, most likely.

Davidski said...

Yes, probably most of it from Greece, but then also from east and south of Greece on top of that.

But it's unlikely that these Roman samples are just from southern Italy, so we're probably talking about much of Italy.

North Italy may have also been affected to some degree, but then moved back up to the Iron Age part of the plot as a result of some Germanic admix.

Wastrel said...

A couple on things on "pre-Germanic".

Germanic has several identifying features, in phonology, lexicon, and morphology.

Early phonological changes are the same as or very similar to those for Italo-Celtic, and aren't remarkable. In fact, they're surprisingly few. The big distinguishing features (Grimm etc) must have happened much later. "pre-Germanic" would have been phonologically conservative (as you'd expect from a peripheral language).

Morphologically, the big thing is the verb. There's been a) simplification of categories, b) analogical regularisation, and c) an entirely new 'weak' past tense. a) and b) are natural processes that happened throughout IE in different ways, because the PIE verbal system was a mess. With the past, what happened was that the stative was adopted as a past, and so verbs that didn't have a stative had to use a periphrastic construction (probably with 'did') that was fossilised as a new suffix. This is unique, but understandable, and wouldn't have seemed weird at the time (like a rural English speaker saying "I did go" (or "I done gone") instead of "I went").

Lexically, about 30% of the vocabulary has no other IE parallels. However, much of this may actually be old words lost in other branches - in the verbs, the oddities are concentrated in the old strong verbs (50% of which are unique to germanic), which would be very weird if they were borrowings, but natural if they're retentions. However, there are certainly some loanwords. Some are shared with Italo-Celtic, while others (particularly maritime terms iirc?) are not. These probably represent one or more non-IE pre-Germanic languages. The only substantial borrowings from IE languages are from Celtic and then from Latin, both much later.

There are also some connexions between Gmc and Balto-Slavic - the spread of the r-passive and m-dative, and perhaps the loss of the old subjunctive (though that may be coincidence) - but these are easily explained as areal features suggesting later contact (with BSl, or with a lost intermediary language).

So 'pre-Germanic' would have been, pre-Grimm, a fairly dull, conservative dialect, with a weird past tense and dialect words - but you find that in English dialects! It's not a "hybrid", or anything that required millennia of isolated development. Similarly of course most of the distinctive features in Celtic are also very late.

We don't know where pre-Germanic was spoken. Nordic Bronze Age is a good guess. But it could also quite feasibly be, say, northern Urnfield (which would have the advantage of making Urnfield the culture of 'Northwestern'. The later replacement of NBA languages with Urnfield ones would be virtually invisible to us - given closeness of IE dialects at that time, it wouldn't require many (if any) actual migrants, and those migrants would be extremely genetically similar to NBA anyway due to recent common ancestry.

Wastrel said...

A quick thing on "Etruscans", incidentally:

We know that Etruscan-speaking people ruled over Italic-speaking people (including the early Latins), and many culturally Etruscan people may therefore have spoken Italic. Given this Etruscan cultural dominance, it's also reasonable to assume that many Etruscan-speakers were themselves Etruscanised Italic people.

So just because a body is found in an Etruscan cultural context, doesn't mean that person spoke Etruscan. And even if they did speak Etruscan, it doesn't mean that they represent, as it were, the ur-Etruscan or 'core Etruscan' tribe.

If the Etruscans being sampled look exactly like the neighbouring/subjugated/Etruscanised Italic people, it may be because those "Etruscans" aren't really Etruscan.

I suppose we could outline five possibilities:

- the Etruscans were always a non-IE steppe-derived people, genetically very similar to the Italics, despite speaking a different language

- Etruscan is a native Italic language (or a later migrant from, say, Liburnia), but early intermixture with Italics (/Illyrians/Venetics/etc) means that the earliest Etruscan culture we know was already genetically the same as the surrounding IE cultures; the only difference is that the local language survived there (the steppe people replaced the entire population but not the language?)

- Etruscan is non-steppe, and there was a native Etruscan culture that came to dominate local steppe-derived cultures. However, the Etruscan ruling class intermarried so extensively with the Italics that 'mature' Etruscan culture was effectively Italic

- genetically distinct non-steppe Etruscans did survive until relatively late, but only represented a minority of culturally Etruscan people, and we just happen not to have found them yet.

I think we need to have a LOT of Etruscan samples, across a span of time and space, before we can isolate a single 'solution' from these options. Particularly because as well as the 'steppe Etruscan' and 'Italian Etruscan' grand theories, there's also a third option (maritime Etruscan, either the traditional 'Trojan' migration from the Aegean, or just a shorter migration from, say, Liburnia/Illyria. Since the presence of Etruscan-related languages in the Aegean demonstrates that they were a navally sophisticated culture at some point in time, we can't rule much out in terms of their origins...)

Desdichado said...

"Tocharian samples from Silk Road sites are on the way.

What if they're packing a lot of Sintashta ancestry and R1a-Z93?"

Then Dragos will have been right, heheh.


Or more likely, they weren't actually "Tocharian" samples at all. We already know that plenty of Sintashta/East Iranic peoples were in the Basin and the surrounding area at a fairly early date. The prehistoric window in which you can get samples that you are sure are Tocharian and not Saka or Sogdian or whatnot is relatively small.

Desdichado said...

So just because a body is found in an Etruscan cultural context, doesn't mean that person spoke Etruscan. And even if they did speak Etruscan, it doesn't mean that they represent, as it were, the ur-Etruscan or 'core Etruscan' tribe.

The same is true for the Romans. Just because we have "Roman" samples doesn't mean that we have Roman samples. It's clear that the Romans were numerically much less numerous than other peoples around them, and that they recognized a clear ethnic division that would be difficult for us to find in our genetic sampling today. It's rather remarkable that the old Patrician families are always described by historical texts as having a phenotype that would be pretty foreign on the streets of Rome today, and that it was already on its way out numerically by the time the Republic turned into the Empire. Appian said that "the city masses are now thoroughly mixed with foreign blood, the freed slave has the same rights as a native-born citizen, and those who are still slaves look no different from their masters." Which was obviously meant to contrast to some past state.

But the point is, that it's a bit difficult to interpret the samples we find. Just because we find a sample in Rome at a certain time period doesn't mean that that sample would have been recognized as "Roman" by the people of Rome.

EurDNA said...

@ Davidski

Yes I agree that genetically Northern BB formed in northern Europe, perhaps the Rhine region, or Mittel-Elbe-Saal.
Still, the details are interesting. For ex ''Kultureller Bruch oder Kontinuität? – Mitteldeutschland im 23. Jh. v. Chr'' describes the situation well for Saxony.

Btw, in the Video, the Hungarian site had females which were brachycranic plano-occiptals; the males were dolichocranic. Shot [1]

Davidski said...

@EurDNA

Rhenish Beakers look like the source of steppe ancestry in all other Beakers, except for some of the steppe ancestry in some Hungarian and Polish Beakers that can be attributed directly to Yamnaya.

And Rhenish Beakers are basically a northwestern extension of German Corded Ware.

https://3.bp.blogspot.com/-38iXPOEhNbg/XEaaaOGlNOI/AAAAAAAAHeI/3l7ZCn0AgyYaI32nspEya1Z76uenmdJ1gCLcBGAs/s1600/Dutch_Beakers_vs_German_Corded_Ware.png

Davidski said...

@EurDNA

Btw, in the Video, the Hungarian site had females which were brachycranic plano-occiptals; the males were dolichocranic.

Yeah, one of these males is the likely first generation Beaker/Yamnaya mix that I mentioned above.

He got his skull from his Z2103 Yamnaya father.

EurDNA said...

@ Davidski
Yes, Hungarian Beaker consists a mixture of 3 or 4 components - northern BB, Hungarian Yamnaya, & local, but also non-local, 'MNE'; and that the link b/w Hungarian BB and other Beaker groups is the archetypcal 'Northern BB'.

Andrzejewski said...

@Desdichado " It's rather remarkable that the old Patrician families are always described by historical texts as having a phenotype that would be pretty foreign on the streets of Rome today"

What phenotype did they have exactly?

Desdichado said...

@Andrzejewski

https://www.theapricity.com/earlson/history/emperors.htm

http://www.unz.com/article/what-race-were-the-greeks-and-romans/

The last article uses a lot of... outdated descriptive language, I guess maybe you can say. But the data is still what it is.

Andrzejewski said...

@Desdichado "@Andrzejewski

https://www.theapricity.com/earlson/history/emperors.htm

http://www.unz.com/article/what-race-were-the-greeks-and-romans/

The last article uses a lot of... outdated descriptive language, I guess maybe you can say. But the data is still what it is."

It only reinforces what I've always tended to think: Caspian-Pontic Steppe Kurganist war bands brought red-hair, blue or green eyes and light complexion into Europe, while the mostly EEF with some WHG admixture natives were largely Mediterranean-looking, not unlike that slaughtered massacred GAC family described in an earlier blog entry.

Andrzejewski said...

Regarding CWC language as the "reconstructed late PIE" in lieu of Yamnaya:

It used to be argued in favor of an EEF substrate within NW IE languages that this particular substrate is common among European IE languages but it is practically absent amongst Indo-Iranian languages (and Tocharian) so it must've survived from one of more non-IE Farmer language.

BUT, with the discovery that Indo-Iranian (Sintashta, Poltavka, Andronovo etc) are largely derived from CORDED WARE and NOT from Yamnaya or other pure Steppe culture, and that Andronovo therefore has at least 20%-30% EEF + some WHG, it completely pulls the rug from underneath this (outdated?) theory...

What do you think?

Gabriel said...

Regarding the Romans: How do modern Central Italians fit into this all?

Blasonario Cremonese said...

@ Wastrel

You wrote a really good post: it's true that Etruscans were dominating Italics IE peoples and also Rome for centuries.

@ Desdichado

You are right! But if we really find that actual Tocharians were R1a, so I think that it would be game over for the R1b-IE link.

" It's rather remarkable that the old Patrician families are always described by historical texts as having a phenotype that would be pretty foreign on the streets of Rome today, and that it was already on its way out numerically by the time the Republic turned into the Empire."

It is also true for original stock Italic plebeian families, like that of Marcus Porcius Cato: he was red haired. But it is also true that many patrician families were actually nobles from neighbours: Etruscans, Sabines, Latins and Romans themselves. In late Republican era, many patricians were proud of their Etruscan paternal line that predated Rome.


@ All

I know that the question of the origin of steppe ancestry in Bell Beaker is wonderful... but I think that it is even more useful to know from where R-L51 (and P312) in Bell Beaker came from... and, frankly, I'm starting to loose my feeling that we will find it in Yamna or something like Yamna subcultures.

capra internetensis said...

@Suyindik

Most of what gets called "leaks" around here are conference presentations, public remarks, and open interviews. We do get a few actual rumours, which are wrong half the time. But it annoys the heck out of me when people call public presentations by the researchers at a major conference "rumours", FFS.

Desdichado said...

You are right! But if we really find that actual Tocharians were R1a, so I think that it would be game over for the R1b-IE link.

Not just any R1a, but a Sintashta/Andronovo-linked R1a-Z93 or related subclade.

The problem is; how do you know any particular corpse that you're taking genetics from is a Tocharian and not a Scythian or other Iranian type? Nobody knows for sure who the Tocharians were. Texts in Tocharian languages are found in certain Tarim Basin cities, but Sogdian and Saka texts are found in those same cities. Nobody knows exactly where the Tocharians came from, other than the reasonably good guess that they might have come with the Afanasevo culture. But let's be clear; the connection between Afanasevo and early proto-Tocharian is merely a good guess.

People presume that some of the tribes may well have been Tocharian; like Wusun or Yuezhi, but they may also not have been, they may have been Iranic/Scythian or even something else entirely and tying historically attested cultures to archaeological material cultures in that region is pretty tough anyway.

So when we say Tocharian samples are coming, I still wonder how we're to know that they're Tocharian and not Scythian.

weure said...

@Davidski I am curios how the early Germanic tribes like the early Anglo-Saxons, Bavarian ME, Collegno, Szolad and Viking/Sigtuna can be placed are they in the Nordic LN-BA slipstream, I guess so....

Ric Hern said...

It will be interesting if Tocharians turn out to be R1a however it will be more interesting if they turn out as R1b L51...

Ryan said...

It seems pretty damned unlikely Tocharians would be L51. Z2103 seems likely but not L51.

FrankN said...

Andrzej: "It used to be argued in favor of an EEF substrate within NW IE languages that this particular substrate is common among European IE languages but it is practically absent amongst Indo-Iranian languages (and Tocharian) so it must've survived from one of more non-IE Farmer language.

BUT, with the discovery that Indo-Iranian (Sintashta, Poltavka, Andronovo etc) are largely derived from CORDED WARE and NOT from Yamnaya or other pure Steppe culture, and that Andronovo therefore has at least 20%-30% EEF + some WHG, it completely pulls the rug from underneath this (outdated?) theory…
"

Only if you think the people who brought farming to Europe spoke the same language, and used the same agricultural terms, as those who brought farming to Iran and India. This may for some extent hold true, see Witzel's "Macro-Caucasian Substrate" he sees at work in some agricultural terms shared by Basques, Caucasians, and Burushaki. But in general, chances are rather slim that ANF, Iran_Neo and Levantine farmers spoke the same language.
In fact I am pretty certain that Levantine PPN people spoke some kind of pre-Semitic, and the pre-Greek Substrate, which is possibly of EEF origin, has been quite throughly screened for any Semitic traces, w/o much success aside from a few cases like OEgypt twr <> (Mino-)taurus "steer, bull".

Davidski said...

@capra internetensis

My comments here about ancient DNA from Italy aren't based on any conference presentations.

Davidski said...

@weure

I am curios how the early Germanic tribes like the early Anglo-Saxons, Bavarian ME, Collegno, Szolad and Viking/Sigtuna can be placed are they in the Nordic LN-BA slipstream, I guess so....

You can plot any ancient or modern samples in the G25 North Euro PCA using the instructions here.

Global25 workshop 3: genes vs geography in Northern Europe

Davidski said...

@Gabriel

Regarding the Romans: How do modern Central Italians fit into this all?

As far as I can tell, modern Central Italians are very similar to the least southern shifted subset of the Romans.

Not sure what this means though, because their position may have been affected by gene flow from the north in the post-Roman period.

Richard Rocca said...

Are there any Copper Age samples from Italy as well?

Davidski said...

@Richard Rocca

Are there any Copper Age samples from Italy as well?

Just the Beakers that have already been published.

Actually, one of the Sicilian Beakers was renamed in the new Harvard data release to Sicily_C_o I4930. That is, from a Beaker to a Copper Age outlier.

Andrzejewski said...

@FrankN "Only if you think the people who brought farming to Europe spoke the same language, and used the same agricultural terms, as those who brought farming to Iran and India. This may for some extent hold true, see Witzel's "Macro-Caucasian Substrate" he sees at work in some agricultural terms shared by Basques, Caucasians, and Burushaki. But in general, chances are rather slim that ANF, Iran_Neo and Levantine farmers spoke the same language.
In fact I am pretty certain that Levantine PPN people spoke some kind of pre-Semitic, and the pre-Greek Substrate, which is possibly of EEF origin, has been quite throughly screened for any Semitic traces, w/o much success aside from a few cases like OEgypt twr <> (Mino-)taurus "steer, bull"."

You entirely miss my point, which is: these terms did NOT come from any substrate language other than PIE itself. And second, reconstructed (purported) late PIE was likely not Yamnaya but Corded Ware-derived.

BTW, where do you think the language isolate Burushaki came from?

Bob Floy said...

I've also wondered whether PIE was a Corded ware phenomenon, rather than Yamnaya, more so after the Iberian revelations.

Ric Hern said...

And that is why I said it will be more interesting if they turn out to be L51. All kinds of scenarios...

«Oldest ‹Older   1 – 200 of 258   Newer› Newest»